Simulation-based defect and repair shape determination for an object for lithography

Abstract

A method for selecting a repair parameter set for a particle-beam-based repair of a defect of an object for lithography, in particular a lithography mask, comprises obtaining measured contour data of the object in surroundings of the defect, determining at least one repair parameter set for a repair of the defect, obtaining at least one simulated aerial image of the object at least partially based on the contour data and the at least one repair parameter set and selecting at least one first repair parameter set from the at least one repair parameter set for performing the repair at least partially based on the simulated aerial image and at least one predetermined target value.

Description

[0001] SIMULATION-BASED DEFECT AND REPAIR SHAPE DETERMINATION FOR AN OBJECT FOR LITHOGRAPHY

[0002] The present invention claims priority to German patent application no. DE io 2024 118 250.4 filed on June 27, 2024 and entitled “Simulationsbasierte Defekt- und Reparaturformbestimmung” which is incorporated herein by reference in its entirety.

[0003] 1. Technical field

[0004] The present invention relates to methods for selecting a repair parameter set for a particle-beam-based repair of a defect of an object for lithography, in particular a lithography mask, for selecting a defect parameter set of a defect of an object for lithography and for particle-beam-based repair of such a defect, and to corresponding computer programs and apparatuses.

[0005] 2. Prior art

[0006] As a consequence of the steady increase in integration density in microelectronics, objects for lithography, in particular lithography masks, must be able to image ever smaller structure elements into a photoresist layer of a wafer. In order to meet these requirements, the exposure wavelength is shifted to ever shorter wavelengths, into the extreme ultraviolet (EUVj wavelength range (10 nm to 15 nm), and to corresponding EUV masks. On account of the ever-decreasing dimensions of the structure or pattern elements, objects for lithography, in particular lithography masks, cannot always be produced without defects that are visible or printable on a wafer. A defect in a lithography mask is therefore duplicated with each exposure process and found on the respective exposed wafer.

[0007] These must therefore be repaired - whenever possible. The repair of defects may e.g. be implemented on the basis of a modified scanning electron microscope, for instance a MeRiT® tool. In conventional methods, a control measurement is provided following the repair: After a defect of a lithography mask has been repaired, the repaired region may be evaluated in order to be able to assess whether the repair process was successful or whether it failed. For this purpose, e.g. wafer printing maybe performed, or an aerial image is recorded using an AIMS™ (Aerial Image Metrology System). One or more aerial images of the repaired region of a mask are recorded by means of the AIMS™. The repaired region of the mask is analysed on the basis of the aerial image or images in order to be able to make a decision as to how to proceed with the mask.

[0008] Zibold, Axel & Scheruebl, Thomas & Menck, Alexander & Brunner, Robert & Greif, Jorn. (2004). Aerial image measurement technique for today's and future 193-nm lithography mask requirements. Proceedings of SPIE - The International Society for Optical Engineering. 10.1117 / 12.568016, for example, describes such an AIMS™ system: The AIMS™ is an optical system for evaluating an object for lithography, in particular a lithography mask. The aerial image recorded by the system is optically equivalent to the image incident on the photoresist of the wafer but magnified and recorded by a CCD camera. Thus, the AIMS™ tool allows a rapid prediction of the printability of critical features on the wafer without the need for performing real wafer printing with the exposure equipment and subsequent SEM measurements of the printed features.

[0009] The AIMS™ systems may operate at different wavelengths: e.g. 248 nm and longer, 193 nm, 157 nm and / or in the EUV wavelength range (e.g. between 10 nm and 15 nm, preferably at approx. 13.5 nm).

[0010] Such methods, as well as those with different control steps, have some disadvantages: Firstly, switching from the repair tool (e.g. a MeRiT® tool) to the control tool (e.g. an AIMS™ system) is time consuming and labour intensive. The two pieces of equipment often are not even in the same country. Secondly, this cycle must be run several times should the control measurement determine that the repair has not been completed satisfactorily.

[0011] The object of the present invention is to rectify these disadvantages at least in part.

[0012] 3. Summary of the invention This object is achieved at least in part by the aspects described herein.

[0013] A first aspect of the invention relates to a method for selecting a repair parameter set for a particle-beam-based repair of a defect of an object for lithography, in particular a lithography mask, comprising: obtaining measured contour data of an object in surroundings of the defect; determining at least one repair parameter set for a repair of the defect; obtaining at least one simulated aerial image of the object at least partially based on the contour data and the at least one repair parameter set; and selecting at least one first repair parameter set from the at least one repair parameter set for performing the repair at least partially based on the simulated aerial image and at least one predetermined target value.

[0014] The inventors have recognized that conventional methods with post-repair monitoring can be improved significantly by virtue of working with a simulated aerial image of the object rather than a control measurement. Instead of performing a control measurement only after the repair has taken place, the obtaining of the simulated aerial image may allow for a selecting of the repair parameter set, on the basis of which the repair should be performed, before the repair takes place. Thus, it is possible to work from the beginning with e.g. optimized repair parameters(s), which may be encompassed by the repair parameter set.

[0015] Previously, the work for determining a repair data set was frequently performed using contour data (e.g. based on contour data such as SEM images) only. However, due to various influencing factors, this often does not lead to sufficiently good results. As explained in e.g.DE io 2022 202 803 Al, the repair material and the (absorber) material of the object for lithography may e.g. have various complex refractive indices and / or other physical properties. Therefore, a repair shape, e.g. expressed by a repair parameter set, may lead to suboptimal functionality of the object for lithography, even though an evaluation on the basis of the contour data would suggest that the corresponding repair shape is ideal. In addition, small, non-printing defects in the vicinity of the defect to be repaired may also lead to the geometrically optimal repair shape based on the contour data not providing the best result in the wafer print or aerial image.

[0016] The insight of the inventors that monitoring should be brought forward, i.e. carried out before the actual repair, and / or that monitoring should be performed on the basis of simulated aerial images based on contours that deviate from measured contour data may therefore make it possible to increase the reliability and speed of the method, and this may significantly reduce costs and outlay in terms of labour.

[0017] Lithography masks described herein may serve as e.g. projection templates for photolithographic structuring in the manufacture of microelectronic circuits or microsystems. They may comprise a substrate, for example made of quartz glass or calcium fluoride, and / or be provided with an absorber layer, e.g. a structured layer, which e.g. may comprise chromium, e.g. on one side of the substrate. It may be an absorptive and / or phase-shifting mask. This means that a variety of different materials may also be used for the absorber layer. The absorber layer comprises in the form of its structure transparent and opaque regions (at least in the spectral range of the exposure used for photolithography). An exemplary structure of an object for lithography is depicted schematically in Fig. 6.

[0018] Typical lithography masks may comprise e.g. binary masks (e.g. chromium masks, 0M0G masks with an opaque MoSi layer on glass, etc.) or phase masks (e.g. chromium-free phase masks, alternating phase masks, rim phase masks, halftone phase masks, tritone phase masks, etc.). Lithography masks maybe tailored to different wavelength ranges. For example, these maybe in the range of "deep ultraviolet light" (DUV light) with wavelengths of 365 nm, 248 nm and / or 193 nm. In other examples, electromagnetic radiation with a wavelength from 10 nm to 20 nm, preferably around 13.5 nm, i.e. in the extreme ultraviolet (EUV) spectral range, maybe used.

[0019] Exemplary (EUV) lithography masks have a reflective multilayer stack that is approx. 300 nm thick, consists of e.g. 20 to too alternating layers of silicon (Si) and molybdenum (Mo) and may optionally be coated with a relatively thin buffer layer, e.g. comprising ruthenium. An absorber material may be applied thereon. The absorber material maybe structured, for example, and contain the pattern information for the wafer exposure process: While the multilayer made of Mo and Si is capable of reflecting the incident light in an exemplary exposure process, the absorber is capable of blocking the reflection (e.g. absorbing the light) and / or reflecting it with a phase offset. Typical thicknesses of the absorber material or the absorber layer are in the range of approx.

[0020] 50 nm to 80 nm.

[0021] Obtaining the measured contour data may e.g. comprise a reception, a generation, a recording, a measurement and / or a retrieval (e.g. from a database). Obtaining the simulated aerial image may e.g. comprise a reception, a generation, a simulation and / or a retrieval (e.g. from a database).

[0022] The surroundings of the defect may encompass the defect at least in part and / or in full.

[0023] The determining of at least one repair parameter set may e.g. be based on the contour data and / or design data (e.g. of at least one pattern element of the absorber layer). The determining maybe understood e.g. as a stipulation. Design data herein refers to information about how the contours of the object for lithography, e.g. the mask, should extend. The contours may e.g. relate to contours of one or more pattern elements (e.g. with absorber material).

[0024] The simulated aerial image of the object, at least partially based on the contour data and the at least one repair parameter set, may be understood to be a simulated aerial image that is based on a composite model of the contour data and the repair parameter set. Hence, the aerial image that is expected to be obtained after the repair can be estimated in advance. Hence, the repair parameter set can be optimized e.g. on the basis of an evaluation in the aerial image space rather than in the contour data space.

[0025] For example, the target value may assist in determining whether a repair parameter set may be judged to be sufficiently suitable for the repair of the defect so that it may be selected. The at least one target value may have at least one value of a reference aerial image, for example. It is emphasized that the simulated aerial image need not be a complete aerial image of the object. Instead, at least a partial simulation, e.g. only at a specific location, of an aerial image that represents surroundings of the defect may be sufficient.

[0026] In some examples, the determining may comprise a determining a plurality of repair parameter sets for the repair of the defect. Further, the obtaining may comprise obtaining a plurality of simulated aerial images of the object at least partially based on the contour data and, from the plurality of repair parameter sets, at least one repair parameter set each. The selecting may comprise selecting at least one first repair parameter set from the plurality of repair parameter sets.

[0027] For example, a plurality of repair parameter sets may thus be available, each with an associated simulated aerial image. On the basis of this, the first repair parameter set which ideally is the most suitable of the plurality of repair data sets, may be selected from the plurality of repair parameter sets in a particularly accurate, reliable and targeted manner.

[0028] A variation of the first aspect relates to a method for selecting a repair parameter set for a particle-beam-based repair of a defect of an object for lithography, in particular a lithography mask, comprising: obtaining measured contour data of the object in surroundings of the defect; determining a plurality of repair parameter sets for a repair of the defect; obtaining a plurality of simulated aerial images of the object at least partially based on the contour data and, from the plurality of repair parameter sets, at least one repair parameter set each; and ascertaining at least one first repair parameter set for performing the repair at least partially based on the plurality of repair parameter sets, the plurality of simulated aerial images and at least one predetermined target value.

[0029] The inventors have recognized that the (pre-)determining the plurality of repair parameter sets allows several advantages: The simulations (often to be minimized in the state of the art) which can now be performed in a cost-effective manner should be accepted, since they involve both reasonable costs and an optimized ascertaining of the (ideal) first repair parameter set. In principle, the first repair parameter set might be included in the plurality of repair parameter sets, or it might not be included in the plurality of repair parameter sets.

[0030] For example (e.g. if the at least one first repair parameter set is not included in the plurality of repair parameter sets), the ascertaining maybe based at least in part on a fitting and / or interpolating using the plurality of repair parameter sets (140a, 140b). For example, a respective metric may be determined for each of the plurality of repair parameter sets. For example, a deviation from a target value may be ascertained as described herein for the respective aerial image obtained for a repair parameter set and maybe used as a metric. For example, the ascertaining maybe based on determining a (local and / or global) minimum or maximum of a function obtained by fitting and / or interpolating. For example, the function may indicate the profile of the metric dependent on the repair parameter set. For example, the minimum or maximum may be found in the first repair parameter set, which is not necessarily included in the plurality of repair parameter sets but maybe obtained e.g. by interpolating between two or more repair parameter sets and / or fitting of a function to two or more repair parameter sets. Thus, the actually optimized first repair parameter set may be advantageously determined, even if it was not necessarily simulated at first. Furthermore, an inaccuracy of the simulations, for example, may thus be compensated for at least in part by fitting or interpolating.

[0031] For example (e.g. if the at least one first repair parameter set is included in the plurality of repair parameter sets), the ascertaining may also comprise selecting the at least one first repair parameter set from the plurality of repair parameter sets. For example, the selecting may also be based on a metric that is associated with the plurality of repair parameter sets. Specifically, the (fitted) metric may for example be a (difference) measure as described herein, e.g. from a target value. For example, the (fitted parameter) may vary with one or more repair parameters of the repair parameter set, and the fitting or interpolation function may comprise one or more local minima (or maxima) and a global minimum (or maximum). For example, the first repair parameter set may be ascertained such that the first repair parameter set corresponds to the repair parameter set that is associated with the (local and / or global) minimum or maximum metric. This maybe a repair parameter set that is included in the plurality of repair parameter sets (or might only be obtained therefrom in other examples). For example, this may be done by selecting from the plurality of repair parameter sets the repair parameter set that is associated with the maximum or minimum metric. Alternatively, the ascertaining maybe based on determining a (local and / or global) minimum or maximum of a function obtained by fitting and / or interpolating. For example, the first parameter set may be the parameter set closest to the (local and / or global) minimum or maximum (for example, even if another one of the plurality of parameter sets is associated with a lower or higher metric). An inaccuracy of the simulations, for example, may thus be compensated for at least in part by fitting or interpolating.

[0032] An exemplary embodiment of the method may furthermore comprise obtaining design data. Further, the determining of the at least one repair parameter set, e.g. a start repair parameter set, may be based at least in part on the contour data and the design data, preferably on comparing the contour data and the design data. That is to say, the repair parameter set may specify one or more parameters that maybe used to modify the contour such that it matches the design data after the repair.

[0033] In essence, a suitable start value for the repair parameter set, for example, may be determined thereby; for example, said start value may be varied from this in order to test other repair parameter sets (e.g. in the form of an associated simulated aerial image) and optionally select these as described herein. The variation may be e.g. systematic in order to find an optimum.

[0034] An example of such comparing is described in relation to Fig. 1.

[0035] In an example, the at least one target value may be based at least in part on a reference aerial image. The at least one target value may comprise e.g. an ideal reference aerial image or one or more pixel value(s) thereof. In a further example, the at least one target value may comprise a region around the ideal reference image, i.e. for example one or more regions(s) of pixel value(s) at certain locations.

[0036] To assess how close a simulated aerial image is to the target value, it is for example possible to check whether the simulated aerial image is in the region around the ideal reference aerial image, i.e., of the order of pixel value(s) at certain locations. This allows the associated repair parameter set to be found to be suitable and / or selected. This can ensure a certain minimum quality of the method.

[0037] In an alternative to that or in addition, a (difference) measure that is based on the simulated aerial image maybe used for such an assessment: The corresponding (difference) measure may e.g. comprise or be based on a difference between at least one pixel value of the simulated aerial image and at least one pixel value of the reference aerial image. For example, the (difference) measure maybe calculated and / or determined using conventional image analysis methods. In an example, the measure may comprise a mean value (or sum) of all deviations (or their absolute values, squares, etc.) between a respective pixel value of the simulated aerial image and the associated pixel value of the reference aerial image in the surroundings of the defect.

[0038] In an alternative to that or in addition, at least one quality parameter, e.g. comprising a critical dimension (CD), a critical dimension uniformity (CDU), a relative critical dimension ACD / CD, a normalized image log-slope (NILS; for example defined as the slope of the intensity logarithm multiplied by the line width) and / or a quantity ascertained from these quality parameters, may be determined for an or each simulated (aerial) image.

[0039] Typically, the first repair parameter set may be selected on the basis of the fact that a (difference) dimension that is as small as possible has been determined therefor and / or that the (difference) dimension and / or the at least one quality parameter is within at least one predetermined range, which may e.g. indicate that a minimum degree of object quality has been reached. In some examples, the at least one quality parameter may also be optimized.

[0040] For example, the selecting may therefore be based on a minimum deviation from the target value.

[0041] For example, in this case it is possible to select the first repair parameter set for which the deviation of the associated simulated aerial image deviates the least from the corresponding target value or reference aerial image. A second aspect of the present invention relates to a method for selecting a 3D defect parameter set of a defect of an object for lithography, in particular a lithography mask, comprising: obtaining measured contour data of the object in surroundings of the defect and of a measured aerial image of the object in surroundings of the defect; determining at least one 3D defect parameter; obtaining at least one simulated aerial image of the object at least partially based on the contour data and the at least one 3D defect parameter set; and selecting at least one first 3D defect parameter set from the at least one 3D defect parameter set at least partially based on the at least one simulated aerial image and the measured aerial image.

[0042] The inventors have recognized that some of the difficulties in conventional defect repair stem from making excessively simplified assumptions about the nature of defects. These typically assume that defects are present either not at all or over the entire depth (e.g. by virtue of defects comprising holes that pass through the whole absorber layer) of e.g. the absorber layer of the object. However, this is frequently not the case:

[0043] Instead, defects are often present as absorber structures that are too low or too high, i.e. situations that cannot be sufficiently described or taken into account under the too simplistic assumptions described herein.

[0044] 3D modelling of a defect can prevent faulty repairs and accelerate and improve corresponding methods e.g. for the repair of the defect.

[0045] The steps of the method according to the first and / or second aspect can be carried out before a repair (e.g. according to the selected repair parameter set).

[0046] In this context, the 3D defect parameter set may comprise at least one extent coordinate in the plane of the object (e.g. x and / or y) and / or an extent coordinate outside of the plane of the object, preferably perpendicular to the plane of the object (e.g. h(x, y)). The at least one extent coordinate may for example define a 2D shape in a plane, preferably in the plane of the object, e.g. a rectangular region or a region in a different geometric shape. Hence, the 3D defect parameter set may represent a three- dimensional model of a defect, which for example may comprise its lateral extent and / or its height profile (e.g. orthogonal to lateral extent). This model can therefore enable an understanding of the present defect that is as detailed as possible and, on the basis thereof, allow a more targeted simulation of an aerial image of the object at least partially based on the contour data and the at least one 3D defect parameter set or optionally (subsequently) allow a corresponding repair of the defect.

[0047] A variation of the second aspect of the invention relates to a method for ascertaining a 3D defect parameter set of a defect of an object for lithography, in particular a lithography mask, comprising: obtaining measured contour data of the object in surroundings of the defect and of a measured aerial image of the object in surroundings of the defect; determining at least one 3D defect parameter set; obtaining at least one simulated aerial image of the object at least partially based on the contour data and the at least one 3D defect parameter set; and ascertaining at least one first 3D defect parameter set at least partially based on the at least one 3D defect parameter set, the at least one simulated aerial image and the measured aerial image.

[0048] For example, the method of the second aspect may be combined with the method of the first aspect. For example, the selecting the at least one first repair parameter set for performing the repair maybe implemented in a manner based at least in part on the selected first 3D defect parameter set.

[0049] In an example, the determining may comprise a determining of a plurality of 3D defect parameter sets, and / or the obtaining may comprise an obtaining of a plurality of simulated aerial images of the object at least partially based on the contour data and, from the plurality of 3D defect parameter sets, at least one 3D defect parameter set each.

[0050] For example, a predetermined plurality may be provided for the method. The larger the plurality, the finer the 3D defect parameter space can be checked, and a suitable selecting of a 3D defect parameter set can be made. Such a selecting may allow the selecting of a first 3D defect parameter set that is particularly close to the actual shape of the defect, which may improve the accuracy of the method.

[0051] In principle, the first 3D defect parameter set might be included in the plurality of 3D defect parameter sets, or it might not be included in the plurality of 3D defect parameter sets. The ascertaining may be carried out as described herein with respect to an ascertaining.

[0052] The method may for example furthermore comprise a determining and / or selecting of at least one repair parameter set for a repair of the defect at least partially based on the (ascertained or selected) first 3D defect parameter set.

[0053] Should the ascertained or selected first 3D defect parameter set be taken into account, the repair maybe planned even more reliably and better. For example, there maybe a case in which a part of the absorber structure is only 20 nm high instead of 60 nm. Then, based on the ascertained or selected first 3D defect parameter set, a repair parameter set that essentially leads to the missing height of approx. 40 nm absorber height being filled up can be easily ascertained or selected. In this example, the actually determined level of the repair may deviate from the intuitively stipulated missing 40 nm, e.g. taking into account further conditions in the surroundings of the defect, but nevertheless result in a substantially optimized repair of the defect.

[0054] In an example, the ascertaining or selecting of the at least one first 3D defect parameter set may be based at least in part on comparing the measured aerial image with a plurality of simulated aerial images, and / or the ascertaining or selecting of the at least one first 3D defect parameter set may be based at least in part on a minimum deviation of the aerial image simulated on the basis of the at least one first 3D defect parameter set from the measured aerial image.

[0055] An ascertaining or selecting as described herein on the basis of a minimum deviation from the target value (for the first aspect of the invention) or on the basis of a minimum deviation of the simulated aerial image from the measured aerial image (for the second aspect of the invention) can be carried out substantially in the same way: Either a predetermined plurality of simulated aerial images may be selected, followed by the selecting of the parameter set for which the smallest deviation is ascertained, and / or simulated aerial images maybe obtained until the respective deviation is below or above a predetermined threshold - in this case, the method can be terminated at this point, and the respective corresponding parameter set maybe ascertained or selected.

[0056] Thus, the first 3D defect parameter set whose associated simulated reference aerial image best matches the actual measured aerial image may be ascertained or selected in a lookup method. This in turn leads to the conclusion that the associated first 3D defect parameter set is the most suitable (out of the plurality considered) for modelling or describing the defect.

[0057] It need not necessarily be the 3D defect parameter set with the minimum deviation that is ascertained or selected.

[0058] In an exemplaiy method, the simulated aerial image may be provided by a trained model. An exemplary trained model is described in DE 10 2018 207882 Al and may be used in the present invention as described herein.

[0059] The use of a trained model offers a number of advantages: Firstly, the greatest amount of time is required in advance (i.e. for example before the use of the trained model according to the methods described herein), specifically during the training of the model. Later, during the use of the model, much less time is required, and so the model can simulate aerial images quickly enough to be able to contribute to the methods described herein substantially in real time or with only small waiting times that do not significantly slow down the repair process.

[0060] In a further exemplary method, the simulated aerial image may be provided or simulated by an algorithm. In order to meet the time requirements of the repair process, the algorithm must e.g. be able to simulate the simulated aerial image of the object in the surroundings of the defect in a sufficiently short period of time. The inventors have found that this time should be less than 10 s, e.g. less than 1 s, less than 0.4 s, and preferably less than 0.3 s. Moreover, the algorithm should achieve a sufficiently accurate simulation. In principle, the method may comprise the simulation of the simulated aerial image.

[0061] In an example, the defect may comprise excess and / or missing mask material, preferably absorber material.

[0062] In an exemplary embodiment of the method, the contour data may be based at least in part on a recording, preferably a scanning particle microscopy (SEM), scanning probe microscopy (AFM) and / or interferometry recording, of the surroundings of the defect.

[0063] Typically, corresponding microscopes are firstly suitable for recording sufficiently high- resolution images, secondly compatible with the planned repair method and thirdly present in any case in conventional repair apparatuses such as conventional MeRiT® tools.

[0064] The contour data may include e.g. an outline of the object for lithography, in particular a lithography mask, preferably the outline of an absorber.

[0065] The outline may thus parameterize the 2D shape of the structure on the object and / or the structure of the object. On the basis of this, defects maybe identified in a first iteration, at least in terms of their type (missing or excess material) and / or position on the object.

[0066] In addition to e.g. microscopic recordings (such as SEM, AFM, interferometry, etc.), i.e. data that contain information about the outline, the contour data may therefore also comprise (e.g. at least partially extracted therefrom) the outline.

[0067] For example, the at least one repair parameter set, preferably the at least one first repair parameter set, may comprise a repair shape and / or a processing parameter, wherein the processing parameter may comprise a parameter for removing excess mask material and / or for depositing missing mask material. This allows the repair parameter set to target both the defect and the specific repair and coordinate these two aspects. The inventors have found that this may have a beneficial effect on the efficiency, accuracy and / or reliability of the method.

[0068] The processing parameter may for example comprise a particle beam parameter such as a wavelength, a focussing, an intensity (e.g. time-varying or time-constant) and / or a position (e.g. relative to the object and / or defect).

[0069] In an example, the repair shape may comprise at least a lateral extent, a height (preferably a height profile) and / or a repair material.

[0070] The repair shape may thus be described in full, and this might facilitate the repair process. The latter may be checked by means of control measurements, for example, so that an optimized repair result can be achieved.

[0071] The repair shape might for example be described by a lateral shape in the x-y plane and / or a height profile expressed as f(x, y). For example, a cuboid repair shape maybe f(x, y) = h0if Xi < x < x2and yx< y < y2are satisfied and otherwise f(x, y) = h0. The repair material may for example specify which material is intended to be deposited, e.g. to augment missing material, or which materials should be used for etching, e.g. various gases (such as an etching gas such as xenon difluoride (XeF2), chlorine (Cl2), oxygen (02), ozone (03), water vapour (H20), hydrogen peroxide (H202), dinitrogen monoxide (N20), nitrogen monoxide (NO), nitrogen dioxide (N02), nitric acid (HN03), nitrosyl chloride (NOCI), ammonia (NH3) and / or sulfur hexafluoride (SFe)).

[0072] Herein, the repair parameter set can define the manner in which a repair is performed. This maybe done, for example, as described herein using the repair shape and / or a processing parameter. By contrast, the 3D defect parameter set may describe the shape of the defect, e.g. by means of an extent coordinate in the plane of the object (e.g. x and / or y) and / or an extent coordinate outside the plane of the object (e.g. h(x, y)). This means that the repair shape (in the repair parameter set) and the extent coordinate(s) (in the 3D defect parameter set) may in general differ from each other, but they may typically be coordinated with each other. In the simplest case, they may substantially correspond. For example, a defect in the form of superfluous mask material within the first extent coordinates of the defect may be repaired or removed by a repair (e.g. etching) according to a first repair parameter set, comprising a first repair shape. In the simplest case in this context, the first extent coordinates and the first repair shape may define the same volume.

[0073] The method may further comprise, for example, an (e.g. particle-beam-based) repair of the defect at least partially based on the at least one first repair parameter set.

[0074] Moreover, the particle-beam-based repair may preferably comprise a particle beam- induced deposition and / or a particle beam-induced etching.

[0075] The method may thus perform an optimized repair, since the at least one repair parameter set as described herein is selected in a particularly advantageous manner, and the repair based on this may be carried out particularly accurately, reliably and purposefully.

[0076] In addition to that or in an alternative, the particle-beam-based repair of the defect may be based at least in part on the contour data, the position of the defect, the simulated aerial image, the first 3D defect parameter set and / or the measured aerial image.

[0077] In an example, the repair may include an addition and / or removal of lithography mask material, preferably absorber material.

[0078] This allows the object to be comprehensively repaired in such a way that it can fulfil its function in lithography substantially without errors (within the framework of accepted error ranges).

[0079] The method may further comprise e.g. a measurement of final contour data (e.g. as described herein) of the lithography mask in surroundings of the repaired defect.

[0080] This may represent an additional control option that is able to ensure that the method has been performed with the required precision and reliability.

[0081] In an example, the method may further comprise simulating a final aerial image of the lithography mask at least partially based on the final contour data. This simulating may also provide an additional control option that is able to ensure that the method has been performed with the required precision and reliability - especially as regards the specific functionality of the object. This can exclude potential sources of error that could still occur when pure contour data are analysed.

[0082] Each parameter set described herein (i.e. repair parameter set or 3D defect parameter set) may comprise one or more parameters.

[0083] A third aspect of the invention relates to a method for the particle-beam-based repair of a defect of an object for lithography, in particular a lithography mask, comprising: a particle-beam-based repair of the defect at least partially based on the at least one first repair parameter set; wherein the at least one first repair parameter set is selected according to the method described herein.

[0084] The repair may therefore be carried out with particular precision, speed and reliability. Essentially, the method has the advantages described herein on account of the selecting of the repair parameter set.

[0085] A further aspect relates to a method for the repair of a defect of an object for lithography, in particular a lithography mask. The method may comprise: obtaining at least one simulated aerial image of the object in surroundings of the defect and / or of at least one reference aerial image; a repair, in particular a particle-beam-based repair, of the defect; obtaining a final aerial image of the object in the surroundings of the defect, preferably following the repair; and comparing the final aerial image with the simulated aerial image and / or the reference aerial image.

[0086] The obtaining the final aerial image may include simulating the final aerial image. In principle, such a method can be used to repair an object for lithography in a closed-loop working cycle, e.g. in an apparatus as described herein (e.g. a MeRiT® tool). In this case, the method may comprise a further repair of the defect, e.g. at least partially based on the comparing. For example, the comparing may show that a defect has not been repaired sufficiently or has been repaired to an excessive extent, and one or more further repairs must be performed in order to repair the defect iteratively and / or in situ until the comparing shows that the defect has been repaired appropriately. Hence, the object need not be removed from the apparatus between the steps of the method (e.g. between two repairs), and this may significantly reduce the amount of time required and increase efficiency. Thus, the repair maybe monitored in situ in the parameter space of interest (e.g. by means of an aerial image comparison).

[0087] Any simulating (e.g. of the reference aerial image, e.g. on the basis of reference contour data, and / or of the final aerial image, e.g. on the basis of final contour data) may be carried out by a computing apparatus of the apparatus used for repair, and / or the apparatus used for repair may comprise an interface to an external computing apparatus (e.g. an external server) that carries out the simulating at least in part and / or provides the apparatus with instructions for the repair on the basis thereof.

[0088] In some exemplary embodiments, the method may further comprise a simulating of the at least one simulated aerial image.

[0089] For example, the simulating may comprise the following steps: simulating a first electromagnetic field for a first portion of the mask and simulating a second electromagnetic field for a second portion of the mask. For example, the simulating of the second electromagnetic field may be based at least in part on the simulating of the first electromagnetic field, and / or the simulating for the first portion and / or the second portion may be based on a modelling of the first portion and / or the second portion as an inhomogeneous medium.

[0090] The subdivision of the simulating into at least two portions can significantly simplify the simulating or at least sub-steps thereof, and this may reduce the required computing time and computing performance. This means that effort and costs can be saved, e.g. throughout the entire method. This allows the provision of an algorithm that meets the requirements in relation to computing time specified herein.

[0091] In essence, the simulating maybe a (computational) propagation of an electromagnetic wave through the first or second portion of the mask (e.g. as described herein). For example, the first portion may correspond to the absorber layer of the mask and the second portion may correspond to the multilayer structure of the mask (or vice versa).

[0092] According to an exemplary embodiment, the method may comprise a simulating of the reflection of the electromagnetic waves within the multilayer structure, and this may comprise the analytical or numerical calculation of reflection coefficients at a boundary between the second portion and the first portion of the photolithography mask, for example wherein the (analytical) reflection coefficients describe the propagation of electromagnetic waves within the multilayer structure. Instead of iteratively propagating the electromagnetic waves through each layer of the multilayer structure, the reflection coefficients at the boundary need only be calculated once to simulate the reflection within the multilayer structure. In this way, it is possible to obtain a highly accurate simulation of the electromagnetic wave propagation within the multilayer with a greatly reduced computing time.

[0093] In an exemplary embodiment, the simulating of the first electromagnetic field and / or the simulating of the second electromagnetic field may be based approximatively on a Helmholtz equation, preferably on a unidirectional Helmholtz equation. For example, the Helmholtz equation may be solved at least in part on the basis of a wave propagation method.

[0094] One advantage of the Helmholtz equation in the simulating of light wave fields is its ability to model a wide range of phenomena and supply precise quantitative predictions regarding the light propagation. Specifically, it may e.g. take into account the frequency and / or (complex) refractive indices in order to supply precise results.

[0095] The unidirectional Helmholtz equation may further simplify the simulating (e.g. in comparison with the conventional Helmholtz equation) and hence e.g. reduce the required computing time and computing performance. This means that effort and costs can be saved, e.g. throughout the entire method.

[0096] The use of the wave propagation method may offer a variety of advantages, as it has, inter alia, e.g. great accuracy and adaptability to the actual conditions present (e.g. in the form of modelling different types of loads and boundary conditions that may occur in real structures) and / or a high efficiency even when simulating large systems (e.g. masks).

[0097] For example, the first portion and / or second portion maybe decomposed into at least two disjoint, homogeneous regions on the basis of characteristic functions, preferably wherein each point of the first portion and / or the second portion may be represented by an affine combination of the characteristic functions.

[0098] This allows the complexity of the simulation to be further reduced, for example in order to save time and costs, without needing to accept excessive loss of accuracy. In particular, in initial experiments by the inventors, the aforementioned affine combinations have proven to be a particularly preferred compromise in terms of accuracy and efficiency.

[0099] For example, the representation (of the first and / or second portion) of the mask by characteristic functions may comprise: an identification of a number of materials of the structures in the first portion, e.g. the absorber pattern or the absorber structure, the photolithography mask, and / or a definition of a characteristic function for each material that indicates the presence of the material at positions (x, y) of the photolithography mask within a subset of an xy-plane (which, for example, substantially corresponds to the mask plane).

[0100] For example, at least one of the characteristic functions may be a bandwidth-limited function, wherein by preference the at least one characteristic function was created by a low pass filter.

[0101] For example, it may further be advantageous to use continuous characteristic functions and / or complex- valued characteristic functions. In this way, the material distribution within the photolithography mask may be described more flexibly, and / or structures with a size below a grid cell of a sampling grid of the simulation can be resolved, and this may lead to approximations with greater accuracy and / or save computing time (in some examples by a factor of approx, too). The inventors have recognized that discretization of the typically used binary characteristic functions for modelling the mask may be problematic. Since binary characteristic functions are discontinuous, the Shannon-Nyquist theorem requires a very high sampling frequency (at least twice the maximum frequency of the signal) and thus requires a very high resolution of a sampling grid used in the simulation. Sampling is inaccurate if the edges of the structures do not correspond to the sampling grid. Moreover, the resolution of the sampling grid depends on the size of the smallest feature. The high resolution of the sampling grid leads in turn to long computing times when creating the aerial image. The characteristic functions are therefore band-limited according to one aspect of the example.

[0102] For example, a band-limited characteristic function may e.g. be a characteristic function f (ui) with angular frequency a>, for which there is a finite frequency )0such that / (&)) = 0 for |u>| > )0. According to the Shannon-Nyquist theorem, firstly, the required sampling frequency of the discretization of a frequency of the discretization of a band-limited characteristic function depends on its maximum frequency. Then again, a given sampling frequency of a discretization of a band-limited characteristic function directly implies the maximum frequency thereof.

[0103] By using band-limited characteristic functions, it is hence possible e.g. to limit the maximum frequency of the characteristic functions. In this way, according to the Shannon-Nyquist theorem, the required sampling frequency is reduced such that a lower resolution sampling grid may be used to discretize the characteristic functions (as in the case of binary characteristic functions). In this way, the calculation times required for the creation or simulation of the aerial image can be reduced. While the sampling grid resolution depends on the smallest feature of the photolithography mask model in the case of binary characteristic functions, the resolution of the sampling grid maybe independent of the feature size of the features in the model, e.g. the design pattern, of the photolithography mask as a result of the use of band-limited characteristic functions.

[0104] For example, the wave propagation method may comprise approximating an analytical Fourier transform by means of a fast Fourier transform. For example, the wave propagation method may comprise approximating an analytical inverse Fourier transform by means of an inverse fast Fourier transform.

[0105] While the wave propagation method cannot be performed with fast Fourier transforms in conventional approaches on account of the dependence of the dispersion relation on the spatial variables (x,y), this can be made possible by the measures described herein. For example, it is possible to use FFTs by rewriting the wave propagation method with the aid of characteristic functions, and the calculation time can thus be reduced.

[0106] In an example, the first portion of the photolithography mask may be decomposed into different materials by virtue of defining for each material a characteristic function that indicates the presence of the material at different locations in the first portion of the photolithography mask, wherein e.g. at least one characteristic function maybe nonbinary.

[0107] A fourth aspect relates to a computer program, wherein the computer program comprises instructions for carrying out the steps of the method described herein.

[0108] A computer program may be written in any desired programming language, including compiled or interpreted languages, and may be provided in any desired form, including as a standalone program or as a module, component, subprogram, or other entity suitable for use in a computer environment.

[0109] A fifth aspect relates to an apparatus for the particle-beam-based repair of a defect of an object for lithography, in particular a lithography mask, comprising: a particle beam source; a means for obtaining measured contour data of an object in surroundings of the defect; a determination means for determining at least one repair parameter set for a repair of the defect; a means for obtaining a simulated aerial image of the object; and a selection means for automatically selecting the at least one repair parameter set for performing a repair at least partially based on the simulated aerial image and at least one predetermined target value. A variation of the fifth aspect relates to an apparatus for the particle-beam-based repair of a defect of an object for lithography, in particular a lithography mask, comprising: a particle beam source; a means for obtaining measured contour data of an object in surroundings of the defect; a determination means for determining a plurality of repair parameter sets for a repair of the defect; a means for obtaining (540) a plurality of simulated aerial images of the object; and a selection means for automatically ascertaining at least one first repair parameter set for performing the repair at least partially based on the plurality of repair parameter sets, the plurality of simulated aerial images and at least one predetermined target value.

[0110] A sixth aspect relates to an apparatus for the particle-beam-based repair of a defect of an object for lithography, in particular a lithography mask, comprising: a particle beam source; a means for obtaining measured contour data of the object in surroundings of the defect and a measured aerial image of the object in surroundings of the defect; a determination means for determining at least one 3D defect parameter; a means for obtaining a simulated aerial image of the object at least partially based on the contour data and the at least one 3D defect parameter; and a selection means for automatically selecting the at least one first 3D defect parameter set at least partially based on the simulated aerial image and the measured aerial image.

[0111] A variation of the sixth aspect relates to an apparatus for the particle-beam-based repair of a defect of an object for lithography, in particular a lithography mask, comprising: a particle beam source; a means for obtaining measured contour data of the object in surroundings of the defect (220c) and a measured aerial image of the object in surroundings of the defect; a determination means for determining at least one 3D defect parameter set; a means for obtaining at least one simulated aerial image of the object at least partially based on the contour data and the at least one 3D defect parameter set; and a selection means for automatically ascertaining at least one first 3D defect parameter set at least partially based on the at least one 3D defect parameter set, the at least one simulated aerial image and the measured aerial image.

[0112] The functions described herein in relation to the various aspects may also be used in conjunction with the variations described in each case.

[0113] In essence, the computer program and the apparatuses may have the advantages described herein in relation to the methods, and vice versa.

[0114] The steps of the methods described herein may also be configured as appropriate means of the apparatuses or instructions of the computer program, and vice versa. Also, various aspects described with reference to one or more specific methods may be applied to other methods.

[0115] In some embodiments, the means may comprise determining, reception, simulation and / or selection means. Each respective means may carry out at least one step of the methods described herein. The respective means may e.g. comprise a data processor and / or a storage apparatus. In some implementations, the apparatus may comprise one or more computers or computer systems that contain one or more data processors which may be configured to execute one or more programs containing a variety of instructions according to the principles described above. Each data processor may contain one or more processor cores, and each processor core may contain logic circuits for processing data. For example, a data processor may contain an arithmetic logic unit (ALU), a control unit, and various registers. Each data processor may contain a cache memory. Each data processor may contain a system-on-chip (SoC) that comprises multiple processor cores, direct-access memory, graphics processing units, one or more controllers, and one or more communication modules. Each data processor may contain millions or billions of transistors.

[0116] The processing of data described herein, such as the determining and / or selecting of repair and 3D defect parameter data sets, may be performed using one or more computers which may contain one or more data processors for data processing, one or more storage apparatuses for storing data and / or one or more computer programs with instructions which, when executed by the one or more computers, cause the one or more computers to perform the processes. The one or more computers may comprise one or more input devices, such as a keyboard, a mouse, a touchpad, and / or a voice input module, and one or more output devices, such as a display and / or a speaker. In some embodiments, the one or more computing devices may comprise digital electronic circuits, computer hardware, firmware, software, or any desired combination of the aforementioned elements. The features relating to the processing of data may be implemented in a computer program product which is tangibly embodied in an information carrier, for example in a machine-readable storage apparatus, for execution by a programmable processor; and method steps may be executed by a programmable processor that runs a program with instructions for performing functions of the described implementations. In an alternative to that or in addition, the program instructions may be encoded on a propagated signal that is an artificially created signal, such as a machine-created electrical, optical or electromagnetic signal that is created in order to encode information for transmission to a suitable receiver apparatus for execution by a programmable processor.

[0117] For example, the computer(s) may be configured so as to be suitable for the execution of a computer program and may for example comprise both general and special microprocessors and one or more processors of any type of digital computer. In general, a processor receives commands and data from a read-only memoiy area or random access memory area, or both. Elements of a computer system include one or more processors for executing commands and one or more storage areas for storing commands and data. In general, a computer system also comprises or is operatively coupled to one or more machine-readable storage media, such as hard disks, magnetic disks, solid-state drives, magneto-optical disks or optical disks, in order to receive data from or transmit data to these, or both. Machine-readable storage media suitable for embodying computer program instructions and computer program data include various forms of non-transitory memory areas, including e.g. semiconductor memory devices such as EPROM, EEPROM, flash memory devices, and solid-state drives; magnetic disks, e.g. internal hard disk drives or removable media; magneto-optical disks; and CD-ROM, DVD-ROM, and / or Blu-ray Discs.

[0118] In some implementations, the above-described methods maybe implemented using software in order to run on one or more mobile computing devices, one or more local computing devices and / or one or more remote computing devices (which may be e.g. cloud computing devices). For example, the software forms procedures in one or more computer programs that are executed on one or more programmed or programmable computer systems, in the mobile computing devices, local computing devices or remote computing systems (which may have different architectures, such as distributed, client / server, grid or cloud systems), each comprising at least one processor, at least one data storage system (including transitory and non-transitory storage and / or storage elements), at least one wired or wireless input device or one wired or wireless port, and at least one wired or wireless output device or one wireless port.

[0119] In some embodiments, the software may be provided on a medium such as CD-ROM, DVD-ROM, Blu-ray Disc, a solid-state drive or a hard disk, which can be read by a general or special programmable computer or is supplied via a network to the computer on which it is executed (encoded in a transmitted signal). The functions maybe executed on a special computer or using special hardware, such as coprocessors. The software maybe implemented in a distributed manner in which different parts of the calculations specified by the software are executed by different computers. Any such computer program is preferably stored on or downloaded to a storage medium or an apparatus (e.g. solid-state memory or solid-state media or magnetic or optical media) which can be read by a programmable computer for general or special purposes, in order to configure and operate the computer when the storage medium or apparatus is read by the computer system in order to perform the methods described herein. The system according to the invention may also be considered to be a computer-readable storage medium that is configured with a computer program, wherein the storage medium configured thus causes a computer system to operate in a specific and predefined manner in order to carry out the functions described herein.

[0120] The embodiments of the present invention described herein and the respective optional features and properties mentioned in this respect should also be understood in the sense that they are disclosed with one another in all combinations. In particular, in the present case, the description of a feature belonging to an embodiment also should not be understood to mean that the feature is essential or indispensable for the function of the embodiment - unless explicitly stated otherwise.

[0121] 4. Description of the figures

[0122] Fig. i shows an exemplary method for selecting a repair parameter set for a particlebeam-based repair of a defect of an object for lithography.

[0123] Fig. 2 shows a faulty repair process, as may occur in the case of unsuitable modelling of a defect.

[0124] Fig. 3a shows an exemplary method for selecting a defect parameter set of a defect of an object for lithography.

[0125] Fig. 3b shows an exemplary comparing of a measured aerial image with a plurality of simulated reference aerial images.

[0126] Fig. 3c shows an exemplary selecting of the at least one first 3D defect parameter set at least partially based on a comparing of the measured aerial image with a plurality of simulated reference aerial images.

[0127] Fig. 4 shows an exemplary method for repairing a defect of an object for lithography, comprising selecting at least one 3D defect parameter.

[0128] Fig. 5 shows a schematic block diagram of some important components of an apparatus for the particle-beam-based repair of a defect of an object for lithography, in particular a lithography mask. Fig. 6 shows a schematic section through an exemplary object for lithography.

[0129] 5. Detailed description of preferred embodiments

[0130] Fig. 1 shows an exemplary method 100 for selecting a repair parameter set 140a, 140b for a particle-beam-based repair of a defect of an object for lithography 120.

[0131] In a first step of the method, measured contour data 110 of the object in the surroundings of the defect are obtained. In the example of Fig. 1, the contour data 110 comprise an SEM image that shows a region of an object comprising two parallel, horizontally extending and spaced apart elements. It is evident that, in the example in Fig. 1, the left-hand element has a rectangular defect on its right-hand side, and the right-hand element has a less strongly recognizable, elongated rectangular defect on its left-hand side.

[0132] In a second step, simulated aerial images 130 are repeatedly obtained. Specifically, an outline 120 is initially determined on the basis of the contour data 110. A first aerial image 130a is simulated or obtained for this outline 120. A first repair shape or an initial repair parameter set 140a, which in the example in Fig. 1 appears to exactly fill the defect in the left-hand element, is determined on the basis of the outline 120 and / or the first simulated aerial image 130a. A second aerial image 130b is simulated or obtained on the basis of the outline 120 and the first repair parameter set 140a. Although one might expect that the one defect could be completely repaired on the basis of the initial repair parameter set 140a, this is not the case, as the second simulated aerial image 130b shows: Both elements deviate in their optical effect from the target result (in Fig. 1 substantially in the form of two parallel rectangular blocks in the aerial image, wherein the edges may be subject to diffraction-limited blurring). Rectangular shapes in aerial images as described herein may be understood as being described in terms of shape and blurring. Consequently, a further repair parameter set 140b is determined, which in the example in Fig. 1 apparently does not ideally fill the defect in the left-hand element. A third aerial image 130c is simulated or obtained on the basis of the outline 120 and the further repair parameter set 140b. Although one might expect that the one defect could be rather overcompensated for on the basis of the second repair parameter set 140a, this is not the case, as the third simulated aerial image 130c shows: The further repair parameter set 140b is adapted to the two defects in such a way that it leads to the desired optical effect (two parallel rectangular blocks in the aerial image) when the two elements interact. In this example, this e.g. allowed the defect in the right-hand part to be rectified without the latter being filled or repaired directly. Instead, the object could be repaired by a single repair (according to repair parameter set 140b). This may be particularly advantageous, for example, if at least one dimension of a defect (e.g. the width of the right-hand defect in Fig. 1) is too narrow to be able to be reliably repaired (and / or optionally even able to be fully characterized).

[0133] In a third step, therefore, the further repair parameter set 140b is selected and, in a fourth step, the object is repaired according to the second repair parameter set 140b, and so it is expected that the repaired object actually has the optical effect according to the third simulated aerial image 130c.

[0134] Fig. 2 shows a faulty repair process 200, as may occur in the case of unsuitable modelling of a defect 220c.

[0135] The left-hand side of Fig. 2 shows an exemplary outline of an object for lithography. This in turn comprises two parallel, horizontally extending and spaced apart elements 220a, 220b. The left-hand element 220a comprises a defect 220c in the form of excess material (to be removed).

[0136] As depicted by the black arrow, the defect 200c in the example in Fig. 2 is repaired under the assumption that the defect has the same height as the object structure. However, the defect 220c in the example in Fig. 2 is in fact higher than the object structure, and so a defect with reduced height 22od remains on the object after the repair. The invention described herein aims to avoid such inadequate repairs and instead directly select suitable repair parameters (e.g. as described herein). For example, the method described herein for selecting a defect parameter set of a defect of an object for lithography may contribute to this end. An exemplary embodiment thereof is described with reference to Fig. 3a: Fig. 3a shows an exemplary method 300 for selecting a defect parameter set 370 of a defect of an object for lithography.

[0137] In the example of Fig. 3a, measured contour data 360 (an SEM image in the example in Fig. 3a) of the object in surroundings of the defect and a measured aerial image 330 of the object in surroundings of the defect are obtained in a first step.

[0138] In a second step, at least one first 3D defect parameter set, which can describe the defect to the best possible extent (see 3D model 380 in Fig. 3a), is selected. The specific selecting of the suitable 3D defect parameter set can be implemented, for example, as shown in Fig. 3b:

[0139] Fig. 3b shows an exemplary comparing 370 of a measured aerial image 372 with a plurality of simulated reference aerial images 373a, 373b, 373c, 373d in order to be able to select the 3D defect parameter set as described herein (e.g. on the basis of the profiles and / or the intensities of the contour data and / or the measured aerial image 330).

[0140] In a first step, an outline 371 is extracted from the contour data obtained. In the example in Fig. 3b, the surroundings of the defect comprise four parallel, horizontally extending and spaced apart elements. On its right-hand side, the second element from the left in this case comprises a defect in the form of excess material to be removed.

[0141] In a second step, the associated measured aerial image 372 is obtained. The influence of the defect can be identified therein at least in part.

[0142] A plurality of 3D defect parameter sets are determined in a third step. Fig. 3b shows four representative examples, in which different defect heights, specifically 57 nm, 40 nm, 20 nm and o nm, are determined for the lateral shape of the defect that was determined on the basis of the outline 371 in the example in Fig. 3b. A respective simulated aerial image 373a, 373b, 373c, 373d is obtained or simulated for each of these four exemplary heights. In a fourth step, each of the simulated aerial images 373a, 373b, 373c, 373d is compared with the measured aerial image 372. The selecting of the at least one 3D defect parameter set may e.g. be based at least in part on this comparing, as follows: A difference image 374a, 374b, 374c, 374b to the measured aerial image 371 is determined for each simulated aerial image 373a, 373b, 373c, 373d. Respective defect regions 375a, 375b, 375c, 375b that are pronounced to different extents may be detected. It is already evident from the exemplary illustrations in Fig. 3b that the difference is least pronounced for the 20 nm simulation, i.e. the difference image 374c. Therein, hardly any signal is recognizable through the image noise, whereas there is a clear signal in the region of the defect in the difference images 374a, 374b, 374b. It may thus be concluded that the actual nature of the object can be described relatively well by the 20 nm simulation.

[0143] The first 3D defect parameter set associated with the aerial image 373c or difference image 374c is therefore selected in a final step (not shown).

[0144] While the description of Fig. 3b has illustrated the selecting process qualitatively, the process can also be illustrated quantitatively, e.g. as described in Fig. 3c: Fig. 3c shows an exemplaiy selecting of the at least one 3D defect parameter set at least partially based on a comparing of the measured aerial image with a plurality of simulated reference aerial images. In Fig. 3c, a measure for the deviation between the measured aerial image and the respective simulated aerial image is plotted as a function of the defect height on which the simulation is based. This measure may comprise or be based on e.g. a difference between at least one pixel value of the measured aerial image 372 and at least one pixel value of one of the simulated aerial images 373a, 373b, 373c, 373d. It is quite evident that the values describe a curve that has a minimum (indicated by an arrow) at approx. 20 nm. Thus, the quantitative analysis in Fig. 3c substantially agrees with the qualitative analysis in Fig. 3b. In other exemplary embodiments, at least one further 3D defect parameter as described herein maybe varied in an alternative or in addition, and the comparing as described herein is based thereon. In a proof-of-concept study, in which the selecting of the at least one 3D defect parameter set could be demonstrated with sufficient accuracy, random noise was added to an aerial image simulated for a 20 nm height defect (with a 57 nm absorber height) in order to artificially obtain the measured aerial image 372. In view of the approach described herein with reference to Figs 3b and 3c, the previously assumed defect height of 20 nm could be obtained again.

[0145] Fig. 4 shows an exemplary method 400 for repairing a defect of an object for lithography, comprising a selecting of at least one 3D defect parameter.

[0146] In this case, (measured) contour data 410 of the object in surroundings of the defect are obtained in a first step. In the example of Fig. 4, the contour data 410 comprise an SEM image that shows a region of an object comprising two parallel, horizontally extending and spaced apart elements. It is evident that the left-hand element has a rectangular defect in the form of excess material in the example in Fig. 4.

[0147] In a second step, an outline 420 is initially determined on the basis of the contour data 410. The outline 420 encompasses two regions assigned to the elements 420a and 420b and the defect on the left-hand element.

[0148] A repair parameter set 440 (e.g. according to the first aspect of the invention) is selected in a third step, and the object is repaired accordingly - in the example in Fig. 4 by e.g. particle-beam-induced etching in the region of the excess material of the defect. In the example of Fig. 4, however, the same problem occurs as described in relation to Fig. 2: As depicted by the black arrow, the defect is repaired in the example in Fig. 4 under the assumption that the defect has the same height as the object structure42oa, 420b. However, the defect 40 in the example in Fig. 4 is in fact higher than the object structure 420a, 420b, and so a defect 42od with reduced height 42od remains on the object after the repair.

[0149] The existence of the defect 42od is expressed in the aerial image 430 of the object recorded in a fourth step. Therein, it is evident that the object does not yet achieve the desired optical effect (two parallel rectangular blocks in the aerial view) after the repair.

[0150] Consequently, together with contour data 460 of the object in surroundings of the defect 42od recorded after the repair, the aerial image 430 can be used to select at least one 3D defect parameter set (as illustrated by the 3D model 480 in Fig. 4, which shows the remaining defect with a low height in comparing with the object structure), for example as described with reference to the second aspect of the invention. As shown in Fig. 4, the 3D defect parameter set may be incorporated into the process 400 in a loop in order to select a further repair parameter set, this time at least partially based on the at least one 3D defect parameter, in a second iteration. A further repair, which is carried out on the basis thereof, can be used for an optimized repair, which may lead to a completely corrected outline 450, as shown e.g. on the right-hand side of Fig. 4, after passing through the further steps of the loop. If, contrary to expectations, this should not be achieved in the second loop of the process 400, the loop maybe run through as many times as necessary until a satisfactory result is achieved.

[0151] Fig. 5 shows a schematic block diagram of some important components of an apparatus 500 for the particle-beam-based repair of a defect of an object for lithography, in particular a lithography mask. The components of the apparatus 500 may also be present individually and / or independently of each other in other exemplary embodiments or not.

[0152] The object 505 maybe any desired microstructured component or structural part. For example, the object 505 may comprise a transmissive or reflective lithography mask or a template for nanoimprint lithography (NIL). Furthermore, the apparatus 500 may be used for analysing and / or repairing for example an integrated circuit (IC), a microscopic system (MEMS, MOEMS) and / or a photonic integrated circuit (PIC). In the examples explained below, the object 505 is a photolithography mask or a NIL stamp.

[0153] The exemplary apparatus 500 in Fig. 5 is a modified scanning electron microscope (SEM). A particle beam source 515 (an electron gun in this example) creates an electron beam 527, which is directed by the beam shaping elements 520 and beam deflecting elements 525 as a focused electron beam 527 onto the object 505 arranged on a sample stage 510.

[0154] The beam shaping elements 520 contain a single-stage condenser system 518. The single-stage condenser system 518 allows e.g. the creation of a focused electron beam 527 on the object 505 with a very small spot diameter on the object 505 (D <2 nm) while simultaneously having a lower kinetic energy of the electrons of the electron beam 527 on the object 505 (E <1 keV). The SEM has a small working distance from the object 505 in order to achieve the small spot diameter on the object 505. The working distance may have dimensions below 3 mm. The low energy electrons allow virtually damage free processing of the object 505 with a very high spatial resolution.

[0155] Moreover, the beam shaping elements 520 may comprise a set of different stops. The beam current of the electron beam 527 is controlled by way of the choice of the appropriate stop.

[0156] The sample stage 510 comprises micromanipulators (not shown in Fig. 5), with the aid of which a defective location of the object 505 can be brought beneath the point of incidence of the electron beam 529 on the object 505. In addition, the sample stage 510 can be displaced in height, i.e. in the beam direction of the electron beam 527, such that the focus of the electron beam 527 comes to rest on the surface of the object 505 (likewise not illustrated in Fig. 5). Furthermore, the sample stage 510 can comprise an apparatus for setting and controlling the temperature, which makes it possible to bring the object 505 to a specified temperature and keep it at this temperature (not indicated in Fig. 5).

[0157] The apparatus 500 in Fig. 5 uses an electron beam 527 from the particle beam source 515 for initiating a local chemical reaction on the object 505. For example, electrons that are incident on the surface of the object 505 cause less damage on the object 505 in comparing with an ion beam for example, even if their kinetic energy varies over a large energy range. However, the apparatus 500 and the method presented here are not restricted to the use of an electron beam 527. Rather, it is possible to use any desired particle beam 527 that is e.g. able to bring about locally a chemical reaction of a precursor gas at the point of incidence 529 of the particle beam 527 on the surface of the object 505. Examples of alternative particle beams are an ion beam, an atomic beam, a molecular beam and / or a photon beam. Furthermore, it is possible to use two or more particle beams in parallel. In particular, it is possible simultaneously to use an electron beam 527 and a photon beam as energy source 515 (not shown in Fig. 5).

[0158] The electron beam 527 can be used for recording an image of the object 505, for instance a lithography mask, in particular of a defective location of the object 505. A detector 530 for detecting backscattered electrons and / or secondary electrons supplies a signal that is proportional to the surface contour and / or composition of the object 505-

[0159] A computer system 540 of the apparatus 500 is able to create an image of the object 505 by scanning or raster-scanning the electron beam 527 over the object 505 with the aid of a control device 545. The control device 545 may be part of the computer system 540, as illustrated in Fig. 5, or maybe embodied as a separate unit (not illustrated in Fig. 5). The computer system 540 may contain algorithms that are realized in hardware, software, firmware or a combination thereof and make it possible to extract an image from the measurement data of the detector 530. A screen of the computer system 540 (not shown in Fig. 5) can display the calculated image. Moreover, the computer system 540 can store the measurement data of the detector 530 and / or the calculated image. Furthermore, the control unit 545 of the computer system 540 may control the particle beam source (e.g. an electron gun 515), the beam imaging and beam shaping elements 520 and 525 and the single-stage condenser system 518. Control signals of the control device 545 may furthermore control the movement of the sample stage 510 by means of the micro-manipulators (not indicated in Fig. 5).

[0160] The computer system 540 (or a separate means, e.g. coupled to the computer system 540) may comprise the means for obtaining measured contour data of an object in surroundings of the defect, the determination means for determining at least one repair parameter set for a repair of the defect, the means for obtaining at least one simulated aerial image of the object, the selection means for automatically selecting at least one first repair parameter set from the at least one repair parameter set for performing the repair at least partially based on the simulated aerial image and at least one predetermined target value, the means for obtaining measured contour data of the object in surroundings of the defect and a measured aerial image of the object in surroundings of the defect, the determination means for determining at least one 3D defect parameter set, the means for obtaining at least one simulated aerial image of the object at least partially based on the contour data and the at least one 3D defect parameter set and / or the selection means for automatically selecting at least one first 3D defect parameter set from the at least one 3D defect parameter set at least partially based on the at least one simulated aerial image and the measured aerial image. The computer system 540 (or a separate means, e.g. coupled to the computer system 540) may e.g. comprise a means for obtaining a final aerial image of the object in the surrounding of the defect (e.g. after a repair) and / or a means for comparing the final aerial image with a reference aerial image or other aerial images (e.g. simulated aerial images, measured aerial images, etc.). Should one or more of the functionalities mentioned herein be carried out by a separate means, e.g. a separate means coupled to the computer system 540, the computer system 540 may e.g. comprise an interface to transmit the information required for the execution of the corresponding functionality (e.g. steps of the methods described herein) and / or receive the results of the functionality and / or information based thereon. For example, a hardware and / or software type device (e.g. in the form of an interface) may be available to this end. For example, a device may be provided to the effect of outputting contour data and / or SEM images and obtaining corresponding quality data and / or repair parameters in response.

[0161] The apparatus 500 may comprise a second detector 535. The second detector 535 may be used to detect the energy distribution of the secondary electrons emitted by the object 505. Hence, the detector 535 allows the composition of the material removed from the object 505 in a local etching process to be analysed. The detector 535 may comprise a SIMS (secondaiy ion mass spectroscopy) detector in an alternative embodiment.

[0162] The apparatus 500 comprises a first container 550 capable of storing a first precursor gas for the purpose of depositing a sacrificial layer. To this end, the first container may store a metal carbonyl for example, for instance molybdenum hexacarbonyl (Mo(CO)6).

[0163] The second supply container 555 may store a second precursor gas which e.g. may be used for producing reference markings. For example, the second precursor gas may comprise tetraethyl orthosilicate (TEOS, Si(OC2H5)4) or chromium hexacarbonyl (Cr(C0)6). The second supply container 555 may e.g. store a second precursor gas in the form of a first etching gas, which allows for the creation of first reference markings, e.g. in the form of depressions in a sacrificial layer. Further, the first etching gas may be used to remove the part of a first sacrificial layer that covers a defect to be repaired. The first etching gas may comprise xenon difluoride (XeF2), in combination with an additive gas, for instance water (vapour) (H20) or nitrogen dioxide (N02). Alternatively, the first etching gas may comprise nitrosyl chloride (NOCI).

[0164] A third storage container 560 may store an additive gas, for example a halide, for instance chlorine (Cl2), a reducing agent, for example ammonia (NH3), or an oxidizing agent, for instance nitrogen dioxide (N02) or water (H20). An additive gas maybe used to assist the deposition of a sacrificial layer and / or assist the generation of reference markings. Moreover, the additive gas of the third gas storage unit 560 may be used to expose the defect after the creation of a first sacrificial layer. It is preferable to use the nitrogen dioxide (N02) additive gas e.g. for depositing material (e.g. in the form of a sacrificial layer) and / or the water (H20) additive gas e.g. for carrying out etching processes.

[0165] In order to process the object 505 arranged on the sample stage 510, i.e., to repair the defect(s) of said object, the apparatus 500 may comprise at least three supply containers for at least a third and a fourth precursor gas. In the exemplary apparatus 500 in Fig. 5, the third precursor gas stored in the fourth container 565 may comprise three different processing gases.

[0166] Further, the fourth supply container 565 may store a third precursor gas in the form of a further deposition gas. The latter is used to deposit missing material on the object 505 with the aid of an electron beam-induced deposition (EBID) process. For example, the material deposited from the fourth supply container should exhibit very good adherence to the object 505 and reproduce the physical and optical properties of the latter to the best possible extent. For example, a main group element alkoxide, for instance TEOS, or a metal carbonyl, for instance molybdenum hexacarbonyl (Mo(CO)6) or chromium hexacarbonyl (Cr(C0)6), may be stored in the fourth supply container 565-

[0167] The fifth supply container 570 may store a fourth precursor gas in the form of a second etching gas. The second etching gas of the fifth supply container 570 may be used to remove from the object 505 excess material on the object 505 with the aid of a local electron beam-induced etching (EBIE) process. Xenon difluoride (XeF2) is an example of a frequently used etching gas. The second etching gas may comprise nitrosyl chloride (N0C1) should the defect comprise a material that is difficult to etch (e.g. such that the defect cannot be removed using the first etching gas).

[0168] The sixth supply container 575 may store a further precursor gas, for instance a further deposition gas or a third etching gas. In a further embodiment, the sixth supply container may store a second additive gas.

[0169] In the exemplaiy apparatus 500 in Fig. 5, each supply container 550, 555, 560, 565, 570, 575 has its own control valve 551, 556, 561, 566, 571, 576, in order to monitor or control the absolute value of the corresponding gas that is provided per unit time, i.e., the gas volumetric flow rate at the location of the incidence of the electron beam 527. The control valves 551, 556, 561, 566, 571 and 576 are controlled and monitored by the control unit 545 of the computer system 540. The partial pressure ratios of the gases provided at the processing location 529 can thus be set over a wide range.

[0170] Furthermore, in the exemplaiy apparatus 500 each supply container 550, 555, 560, 565, 570, 575 has its own gas feed line system 552, 557, 562, 567, 572, 577, which ends with a nozzle in the vicinity of the point of incidence of the electron beam 527 on the object 505. In an alternative embodiment (not represented in Fig. 5), a gas feed line system is used to bring a plurality or all of the processing gases to the surface of the object 505 in a common stream.

[0171] In the example illustrated in Fig. 5, the valves 551, 556, 561, 566, 571, 576 are arranged in the vicinity of the corresponding containers 550, 555, 560, 565, 570, 575. In an alternative arrangement, the control valves 551, 556, 561, 566, 571, 576 maybe installed in the vicinity of the corresponding nozzles (not shown in Fig. 5). Unlike the illustration shown in Fig. 5 and without preference at the present time, it is also possible to provide one or more of the gases stored in the containers 550, 555, 560, 565, 570, 575 non- directionally in the lower part of the vacuum chamber 502 of the apparatus 500. In this case, it is necessary for the apparatus 500 to incorporate a stop (not illustrated in Fig. 5) between the lower reaction space 502 and the upper part of the apparatus 500, which provides the electron beam 527, in order to prevent an excessively low vacuum in the upper part of the apparatus 500. Each of the supply containers 550, 555, 560, 565, 570 and 575 may have its own temperature setting element and control element that enables both cooling and heating of the corresponding supply containers. This makes it possible to store and provide the deposition gases, the additive gases and the etching gases at the respective optimum temperature (not shown in Fig. 5). Further, the vapour pressure of the precursor gas or gases can be regulated by way of the temperature in the supply container or containers in the case of solid or liquid precursors. The gas volumetric flow rate of gaseous precursors can be controlled with the aid of a mass flow controller (MFC).

[0172] Furthermore, each feeder system 552, 557, 562, 567, 572 and 577 may comprise its own temperature setting element and temperature control element in order to provide all the process gases at their optimum processing temperature at the point of incidence of the electron beam 527 on the object 505 (likewise not indicated in Fig. 5). The control device 545 of the computer system 540 may control the temperature setting elements and the temperature control elements both of the supply containers 550, 555, 560, 565, 570, 575 and of the gas feed line systems 552, 557, 562, 567, 572, 577 and regulate the gas volumetric flow rate through the MFC or MFCs.

[0173] The apparatus 500 in Fig. 5 may comprise a pump system for creating and maintaining a required vacuum in the reaction chamber 502 (not shown in Fig. 5). A residual gas pressure of e.g. <io-6mbar is achieved in the reaction chamber 502 of the apparatus 500 when the control valves 551, 556, 561, 566, 571, 576 are closed. The pump system may comprise separate pump systems both for the upper part of the apparatus 500 for providing the electron beam 527 and for the lower part comprising the reaction chamber 502 with the sample stage 510 with the object 505. Further, the apparatus 500 may comprise a suction extraction apparatus in the vicinity of the processing point 529 of the electron beam 527 in order to define a defined local pressure condition at the surface of the object 505 (not illustrated in Fig. 5). The use of an additional suction extraction apparatus may largely prevent one or more volatile reaction products of the deposition gases, additive gases and the etching gases which are not needed in the local particle beam-induced processes from depositing on the object 505 and / or in the reaction chamber 502. The functions of the pump system or systems and of the additional suction extraction apparatus may likewise be controlled and / or monitored by the control device 545 of the computer system 540. The control device 545, the computer system 540 or a dedicated component of the computer system 540 may ascertain the size of one or more reference markings for an identified defect. The size of a reference marking comprises the determining of both its area and its height. Further, the control device 545, the computer system 540 or a specific component of the computer system 540 may ascertain a scanning region of the electron beam 527 that is used to scan the position of the reference marking(s).

[0174] Further, with knowledge of the material composition of the object 505, the control device 545 is able to select a precursor gas for depositing one or more sacrificial layers 540. Moreover, the control device 545 may select one or more precursor gases and optionally an additive gas for depositing one or more reference markings on the sacrificial layers 540. By choosing suitable material compositions of the sacrificial layer(s) 540 and of the reference markings, it is possible to optimize the visibility of the reference markings against the background of the sacrificial layer(s) 540.

[0175] Fig. 6 shows a schematic section through an exemplary object 600 for lithography:

[0176] Fig. 6 shows a schematic section through a part of an object 600 for lithography, in particular a lithography mask, e.g. for the EUV wavelength range. A many layer structure 620 or multilayer structure 620 is applied to the substrate 610 thereof. For example, the multilayer structure may have 20 to 60 alternating layers made of molybdenum (Mo) and Silicon (Si). An absorber layer in the form of two structure or pattern elements 630a and 630b is attached to or deposited on the multilayer structure 620. In some examples, a buffer layer (not shown) may also be applied between the multilayer structure 620 and the pattern elements 630a and 630b, optionally with an optional capping layer between the multilayer structure 620 and buffer layer (not shown). Finally, a surface layer, e.g. in the form of a passivation layer, may be located on the pattern elements 630a and 630b.

[0177] Further examples:

[0178] 1. Method (too) for selecting a repair parameter set for a particle-beam-based repair of a defect (220c) of an object (600) for lithography, in particular a lithography mask, comprising: obtaining measured contour data (no) of the object (6oo) in surroundings of the defect (220c); determining at least one repair parameter set (140a, 140b) for a repair of the defect (220c); obtaining at least one simulated aerial image (130) of the object (600) at least partially based on the contour data (no) and the at least one repair parameter set (140a, 140b); and selecting at least one first repair parameter set (140b) from the at least one repair parameter set (140a, 140b) for performing the repair at least partially based on the simulated aerial image (130) and at least one predetermined target value.

[0179] 2. Method (100) according to Example 1, wherein the determining comprises determining a plurality of repair parameter sets (140a, 140b) for the repair of the defect (220c); wherein the obtaining comprises obtaining a plurality of simulated aerial images (130) of the object (600) at least partially based on the contour data (110) and, from the plurality of repair parameter sets (140a, 140b), at least one repair parameter set (140a, 140b) each; and wherein the selecting comprises selecting the at least one first repair parameter set (140b) from the plurality of repair parameter sets (140a, 140b).

[0180] 3. Method (too) according to Example 1 or 2, furthermore comprising obtaining design data, wherein the determining of the at least one repair parameter set (140a, 140b) is based at least in part on the contour data (110) and the design data.

[0181] 4. Method (too) according to any of the preceding examples, wherein the at least one predetermined target value is based at least in part on a reference aerial image and / or comprises the reference aerial image.

[0182] 5. Method (too) according to any of the preceding examples, wherein the selecting is based on a minimum deviation from the target value.

[0183] 6. Method (300) for selecting a 3D defect parameter set (370) of a defect (220c) of an object (600) for lithography, in particular a lithography mask, comprising: obtaining measured contour data (360) of the object (600) in surroundings of the defect (220c) and of a measured aerial image (330) of the object (600) in surroundings of the defect (220c); determining at least one 3D defect parameter set (370); obtaining at least one simulated aerial image (373a, 373b, 373c, 373d) of the object (600) at least partially based on the contour data (360) and the at least one 3D defect parameter set (370); and selecting at least one first 3D defect parameter set (370) from the at least one 3D defect parameter set (370) at least partially based on the at least one simulated aerial image (373a, 373b, 373c, 373d) and the measured aerial image (330).

[0184] 7. Method (300) according to Example 6, further comprising determining and / or selecting at least one first repair parameter set (140b) for a repair of the defect (220c) at least partially based on the first 3D defect parameter set (370).

[0185] 8. Method (300) according to either of Examples 6 and 7, wherein the selecting of the at least one first 3D defect parameter set (370) is based at least in part on comparing the measured aerial image (330) with a plurality of simulated aerial images (373a, 373b, 373c, 373d); and / or wherein the selecting of the at least one first 3D defect parameter set (370) is based at least in part on a minimum deviation of the aerial image (373a, 373b, 373c, 373d), simulated on the basis of the at least one first 3D defect parameter set (370), from the measured aerial image (330).

[0186] 9. Method (too, 300) according to any of the preceding examples, wherein the at least one simulated aerial image (130, 373a, 373b, 373c, 373d) is provided by a trained model.

[0187] 10. Method (too, 300) according to any of the preceding examples, wherein the defect (220c) comprises excess and / or missing mask material, preferably absorber material (630a, 630b).

[0188] 11. Method (too, 300) according to any of the preceding examples, wherein the contour data are based at least in part on a recording, preferably a scanning particle microscopy, scanning probe microscopy and / or interferometry recording, of the surroundings of the defect (220c).

[0189] 12. Method (too, 300) according to any of the preceding examples, wherein the contour data comprise an outline of the object (600) for lithography, in particular of a lithography mask, preferably of an absorber material (630a, 630b). 13- Method (100, 300) according to any of Examples 1-5, 7, or 8-12, insofar as referring back to Example 7, wherein the at least one first repair parameter set (140b) comprises a repair shape and / or a processing parameter, wherein the processing parameter comprises a parameter for removing excess mask material and / or for depositing missing mask material.

[0190] 14. Method (100, 300) according to Example 13, wherein the repair shape comprises at least a lateral extent, a height, preferably a height profile, and / or a repair material.

[0191] 15. Method (100, 300) according to any of Examples 1-5, 7, or 8-14, insofar as referring back to Example 7, further comprising: a particle-beam-based repair of the defect (220c) at least partially based on the at least one first repair parameter set (140b), wherein the particle-beam-based repair preferably comprises a particle beam-induced deposition and / or a particle beam-induced etching.

[0192] 16. Method (too, 300) according to any of Examples 1-5, 7, or 8-14, insofar as referring back to Example 7, or 15, wherein the repair includes adding and / or removing lithography mask material, preferably absorber material (630a, 630b).

[0193] 17. Method (too, 300) according to Example 15 or 16, further comprising a measurement of final contour data of the lithography mask in surroundings of the repaired defect (220c).

[0194] 18. Method (too, 300) according to Example 17, further comprising simulating a final aerial image of the lithography mask at least partially based on the final contour data.

[0195] 19. Method according to Example 18, further comprising comparing the final aerial image with the simulated aerial image (130) for the first repair parameter set (140b) and / or a reference aerial image.

[0196] 20. Method according to Example 18, referring back to any of Examples 6-17, further comprising comparing the final aerial image with the measured aerial image (330) and / or a reference aerial image.

[0197] 21. Method (too, 300) for the particle-beam-based repair of a defect (220c) of an object (600) for lithography, in particular a lithography mask, comprising: a particle-beam-based repair of the defect (220c) at least partially based on at least one first repair parameter set (140b); wherein the at least one first repair parameter set (140b) is selected according to the method (100, 300) of any of Examples 1-5, 7, or 8-14, insofar as referring back to Example 7, or 15-20. Method for repairing a defect (220c) of an object (600) for lithography, in particular a lithography mask, comprising: obtaining at least one simulated aerial image (130) of the object (600) in surroundings of the defect (220c) and / or of a reference aerial image; repairing, in particular a particle-beam-based repairing, of the defect (220c); obtaining a final aerial image of the object (600) in the surroundings of the defect (220c); and comparing the final aerial image with the simulated aerial image (330) and / or the reference aerial image. Method (100, 300) according to any of the preceding examples, further comprising simulating the at least one simulated aerial image (130, 373a, 373b, 373c, 373d) , wherein the simulating comprises: simulating a first electromagnetic field for a first portion of the mask; simulating a second electromagnetic field for a second portion of the mask; wherein the simulating the second electromagnetic field is based at least in part on the simulating the first electromagnetic field; wherein the simulating for the first portion and / or the second portion is based on a modelling of the first portion and / or the second portion as an inhomogeneous medium. Method according to Example 23, wherein the simulating of the first electromagnetic field and / or the simulating of the second electromagnetic field is based approximatively on a Helmholtz equation, preferably on a unidirectional Helmholtz equation; and wherein during the simulating, the Helmholtz equation is solved based at least in part on a wave propagation method. Method for compensation according to Example 23 or 24, wherein the first portion and / or the second portion is decomposed into at least two disjoint, homogeneous regions on the basis of characteristic functions, preferably wherein each point of the first portion and / or of the second portion may be represented by an affine combination of the characteristic functions. Method for compensation according to Example 25, wherein at least one of the characteristic functions is a bandwidth-limited function, preferably wherein the at least one characteristic function was created by a low pass filter. Method for compensation according to any of Examples 24, or 25, 26, insofar as referring back to Example 24, wherein the wave propagation method comprises approximating an analytic Fourier transform by way of a fast Fourier transform; and / or wherein the wave propagation method comprises an approximating an analytic inverse Fourier transform byway of an inverse fast Fourier transform. Computer program comprising instructions for executing the steps of the method according to any of the preceding examples. Apparatus for the particle-beam-based repair of a defect (220c) of an object (600) for lithography, in particular a lithography mask, comprising: a particle beam source (515); a means for obtaining (540) measured contour data of an object (600) in surroundings of the defect (220c); a determination means (540) for determining at least one of repair parameter set (140a, 140b) for a repair of the defect (220c); a means for obtaining (540) at least one simulated aerial image (130) of the object (600); and a selection means (540) for automatically selecting at least one first repair parameter set (140b) from the at least one repair parameter set (140a, 140b) for performing the repair at least partially based on the simulated aerial image (130) and at least one predetermined target value. Apparatus for the particle-beam-based repair of a defect (220c) of an object (600) for lithography, in particular a lithography mask, comprising: a particle beam source; a means for obtaining (540) measured contour data (360) of the object (600) in surroundings of the defect (220c) and of a measured aerial image (330) of the object (600) in surroundings of the defect (220c); a determination means (540) for determining at least one 3D defect parameter set (370); a means for obtaining (540) at least one simulated aerial image (373a, 373b, 373c, 373d) of the object (600) at least partially based on the contour data (110) and the at least one 3D defect parameter set (370); and a selection means (540) for automatically selecting at least one first 3D defect parameter set (370) from the at least one 3D defect parameter set

[0198] (370) at least partially based on the at least one simulated aerial image (373a? 373b, 373c, 373d) and the measured aerial image (360).

Claims

Claims1. Method (300) for selecting a 3D defect parameter set (370) of a defect (220c) of an object (600) for lithography, in particular a lithography mask, comprising: obtaining measured contour data (360) of the object (600) in surroundings of the defect (220c) and of a measured aerial image (330) of the object (600) in surroundings of the defect (220c); determining at least one 3D defect parameter set (370); obtaining at least one simulated aerial image (373a, 373b, 373c, 373d) of the object (600) at least partially based on the contour data (360) and the at least one 3D defect parameter set (370); and ascertaining at least one first 3D defect parameter set (370) at least partially based on the at least one 3D defect parameter set (370), the at least one simulated aerial image (373a, 373b, 373c, 373d) and the measured aerial image (330).

2. Method (300) according to Claim 1, further comprising determining and / or selecting at least one first repair parameter set (140b) for a repair of the defect (220c) at least partially based on the first 3D defect parameter set (370).

3. Method (300) according to either of Claims 1 and 2, wherein the selecting of the at least one first 3D defect parameter set (370) is based at least in part on comparing of the measured aerial image (330) with a plurality of simulated aerial images (373a, 373b, 373c, 373d); and / or wherein the selecting the at least one first 3D defect parameter set (370) is based at least in part on a minimum deviation of the aerial image (373a, 373b, 373c, 373d), simulated on the basis of the at least one first 3D defect parameter set (370), from the measured aerial image (330).

4. Method (too) for selecting a repair parameter set for a particle-beam-based repair of a defect (220c) of an object (600) for lithography, in particular a lithography mask, comprising:obtaining measured contour data (no) of the object (6oo) in surroundings of the defect (220c); determining a plurality of repair parameter sets (140a, 140b) for a repair of the defect (220c); obtaining a plurality of simulated aerial images (130) of the object (600) at least partially based on the contour data (no) and, from the plurality of repair parameter sets (140a, 140b), at least one repair parameter set (140a, 140b) each; and ascertaining at least one first repair parameter set (140b) for performing the repair at least partially based on the plurality of repair parameter sets (140a, 140b), the plurality of simulated aerial images (130) and at least one predetermined target value.

5. Method (too) according to Claim 4, wherein the ascertaining comprises selecting the at least one first repair parameter set (140b) from the plurality of repair parameter sets (140a, 140b) and / or wherein the ascertaining is based at least in part on fitting and / or interpolating using the plurality of repair parameter sets (140a, 140b).

6. Method (too) according to Claim 4 or 5, furthermore comprising obtaining design data, wherein the determining the at least one repair parameter set (140a, 140b) is based at least in part on the contour data (110) and the design data.

7. Method (too) according to any of Claims 4-6, wherein the at least one predetermined target value is based at least in part on a reference aerial image and / or comprises the reference aerial image.

8. Method (too) according to any of Claims 4-7, wherein the ascertaining is based on a minimum deviation from the target value.

9. Method (too, 300) according to any of the preceding claims, wherein the at least one simulated aerial image (130, 373a, 373b, 373c, 373d) is provided by a trained model.

10. Method (100, 300) according to any of the preceding claims, wherein the defect (220c) comprises excess and / or missing mask material, preferably absorber material (630a, 630b).

11. Method (100, 300) according to any of the preceding claims, wherein the contour data are based at least in part on a recording, preferably a scanning particle microscopy, scanning probe microscopy and / or interferometry recording, of the surroundings of the defect (220c).

12. Method (100, 300) according to any of the preceding claims, wherein the contour data comprise an outline of the object (600) for lithography, in particular of a lithography mask, preferably of an absorber material (630a, 630b).

13. Method (100, 300) according to any of Claims 2, 4-8 or, insofar as referring back to Claim 2, according to any of Claims 3 or 9-12, wherein the at least one first repair parameter set (140b) comprises a repair shape and / or a processing parameter, wherein the processing parameter comprises a parameter for removing excess mask material and / or for depositing missing mask material.

14. Method (100, 300) according to Claim 13, wherein the repair shape comprises at least a lateral extent, a height, preferably a height profile, and / or a repair material.

15. Method (too, 300) according to any of Claims 2, 4-8, or, insofar as referring back to Claim 2, according to any of Claims 3 or 9-14, further comprising: a particle-beam-based repair of the defect (220c) at least partially based on the at least one first repair parameter set (140b), wherein the particle-beam-based repair preferably comprises a particle beam-induced deposition and / or a particle beam-induced etching.

16. Method (too, 300) according to any of Claims 2, 4-8, or, insofar as referring back to Claim 2, according to any of Claims 3 or 9-15, wherein the repaircomprises an adding and / or removing lithography mask material, preferably absorber material (630a, 630b).

17. Method (100, 300) according to Claim 15 or 16, further comprising measuring final contour data of the lithography mask in surroundings of the repaired defect (220c).

18. Method (100, 300) according to Claim 17, further comprising simulating a final aerial image of the lithography mask at least partially based on the final contour data.

19. Method according to Claim 18, further comprising comparing the final aerial image with the simulated aerial image (130) for the first repair parameter set (140b) and / or a reference aerial image.

20. Method according to Claim 18, referring back to any of Claims 6-17, further comprising comparing the final aerial image with the measured aerial image (330) and / or a reference aerial image.

21. Method (too, 300) for the particle-beam-based repair of a defect (220c) of an object (600) for lithography, in particular a lithography mask, comprising: particle-beam-based repairing of the defect (220c) at least partially based on at least one first repair parameter set (140b); wherein the at least one first repair parameter set (140b) is selected according to the method (too, 300) of Claims 2, 4-8, or, insofar as referring back to Claim 2, according to any of Claims 3 or 9-20.

22. Method for repairing a defect (220c) of an object (600) for lithography, in particular a lithography mask, comprising: obtaining at least one simulated aerial image (130) of the object (600) in surroundings of the defect (220c) and / or of a reference aerial image; repairing, in particular a particle-beam-based repairing, of the defect (220c);obtaining a final aerial image of the object (6oo) in the surroundings of the defect (220c); and comparing the final aerial image with the simulated aerial image (330) and / or the reference aerial image.

23. Method (100, 300) according to any of the preceding claims, further comprising simulating the at least one simulated aerial image (130, 373a, 373b, 373c, 373d), wherein the simulating comprises: simulating a first electromagnetic field for a first portion of the mask; simulating a second electromagnetic field for a second portion of the mask; wherein the simulating of the second electromagnetic field is based at least in part on the simulating of the first electromagnetic field; wherein the simulating for the first portion and / or the second portion is based on a modelling of the first portion and / or the second portion as an inhomogeneous medium.

24. Method according to Claim 23, wherein the simulating of the first electromagnetic field and / or the simulating of the second electromagnetic field is based approximatively on a Helmholtz equation, preferably on a unidirectional Helmholtz equation; and wherein during the simulating, the Helmholtz equation is solved based at least in part on a wave propagation method.

25. Method for compensation according to Claim 23 or 24, wherein the first portion and / or the second portion is decomposed into at least two disjoint, homogeneous regions on the basis of characteristic functions, preferably wherein each point of the first portion and / or of the second portion may be represented by an affine combination of the characteristic functions.

26. Method for compensation according to Claim 25, wherein at least one of the characteristic functions is a bandwidth-limited function, preferably wherein the at least one characteristic function was created by a low pass filter.

27. Method for compensation according to any of Claims 24, or 25, 26, insofar as referring back to Claim 24, wherein the wave propagation method comprises an approximating of an analytic Fourier transform by way of a fast Fourier transform; and / or wherein the wave propagation method comprises an approximating of an analytic inverse Fourier transform by way of an inverse fast Fourier transform.

28. Computer program comprising instructions for executing the steps of the method according to any of the preceding claims.

29. Apparatus for the particle-beam-based repair of a defect (220c) of an object (600) for lithography, in particular a lithography mask, comprising: a particle beam source (515); a means for obtaining (540) measured contour data of an object (600) in surroundings of the defect (220c); a determination means (540) for determining a plurality of repair parameter sets (140a, 140b) for a repair of the defect (220c); a means for obtaining (540) a plurality of simulated aerial images (130) of the object (600); and a selection means (540) for automatically ascertaining at least one first repair parameter set (140b) for performing the repair at least partially based on the plurality of repair parameter sets (140a, 140b), the plurality of simulated aerial images (130) and at least one predetermined target value.

30. Apparatus for the particle-beam-based repair of a defect (220c) of an object (600) for lithography, in particular a lithography mask, comprising: a particle beam source; a means for obtaining (540) measured contour data (360) of the object (600) in surroundings of the defect (220c) and of a measured aerial image (330) of the object (600) in surroundings of the defect (220c); a determination means (540) for determining at least one 3D defect parameter set (370);a means for obtaining (540) at least one simulated aerial image (373a, 373b, 373c, 373d) of the object (600) at least partially based on the contour data (110) and the at least one 3D defect parameter set (370); and a selection means (540) for automatically ascertaining at least one first 3D defect parameter set (370) at least partially based on the at least one 3D defect parameter set (370), the at least one simulated aerial image (373a, 373b, 373c, 373d) and the measured aerial image (360).

Images