Method and apparatus for calibrating motion relative to a photomask - Patents.com

By arranging correction marks in geometrically distinct groups on lithographic masks, the method improves the efficiency and accuracy of mask repair processes, particularly for large area defects, by reducing wear and maintaining consistent calibration.

JP2024543141AActive Publication Date: 2024-11-19CARL ZEISS SMT GMBH
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
JP2024531172
Authority / Receiving Office
JP · JP
Patent Type
Applications
Current Assignee / Owner
Priority Date
2021-11-23
Filing Date
2022-11-22
Publication Date
2024-11-19
Estimated Expiration
2042-11-22

AI Technical Summary

Technical Problem

Existing methods for repairing large area defects on lithographic masks are inefficient due to frequent scanning of fiducial marks, leading to wear and reduced accuracy, especially with increased integration density in semiconductor manufacturing.

Method used

The method involves generating correction marks in geometrically arranged groups on the mask, with inner group separations smaller than outer group separations, allowing for optimized calibration and reduced wear through targeted selection of correction marks for calibration.

Benefits of technology

This approach enhances the accuracy and efficiency of mask repair by minimizing mark degradation and maintaining consistent calibration over extended processing times, even for large area defects.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 2024543141000001_ABST
    Figure 2024543141000001_ABST
Patent Text Reader

Abstract

The present invention relates to a method and apparatus for calibrating operations on a mask. A method for generating correction marks on a lithographic object using a particle beam, in particular for calibrating operations, includes (a.) generating a first group of correction marks, (b.) generating a second group of correction marks, and (c.) a separation of the correction marks in the first group and in the second group is smaller than a separation between the correction marks from the first group and the correction marks from the second group.
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] This patent application claims priority to German patent application DE 10 2021 213 163.8 entitled “Verfahren und Vorrichtung zur Kalibrierung eines Arbeitsvorgangs auf einer Photomaske”, filed with the German Patent Office on November 23, 2021. German patent application DE 10 2021 213 163.8 is incorporated by reference in its entirety into this patent application.

[0002] The present invention relates to a method and an apparatus for calibrating a movement relative to a mask. In particular, the present invention relates to a method for generating correction marks on a lithographic object, a method for calibrating the movement, and a corresponding apparatus and computer program for carrying out the method. [Background technology]

[0003] In the semiconductor industry, smaller and smaller structures are formed on wafers in order to ensure an increased integration density. To form the structures, lithographic methods are used here in particular to image said structures onto the wafer. By way of example, lithographic methods can include photolithography, UV lithography, DUV lithography, EUV lithography, X-ray lithography, nanoimprint lithography, etc. In this process, lithography usually makes use of a mask (e.g. photomask, exposure mask, reticle, stamp in the case of nanoimprint lithography, etc.) which contains a pattern for imaging the desired structures onto the wafer.

[0004] As integration density increases, the demands on mask production also increase (e.g., as a result of the concomitant reduction in structure dimensions on the mask or as a result of higher material requirements in lithography). As a result, the mask production process becomes more and more complex, more and more time-consuming, and more and more expensive, and it is not always possible to avoid mask errors (e.g., defects). Therefore, mask errors are usually repaired or repaired in further processing steps.

[0005] As an example, mask errors can be repaired by a particle beam based process, where the corresponding operations can include processing in a writing field and / or recording an image in an image field, in which a particle beam is typically scanned over a given writing or image field in a targeted manner (e.g., along a rectangular pixel raster).

[0006] Because the mask is typically an electrically insulating specimen, scanning the mask with charged particles (e.g., electrons from an electron beam, ions from an ion beam) can cause electrostatic charging of the mask, which can cause the particle beam to unintentionally deviate from its intended point of incidence. This effect is called particle beam position drift, or drift, and other mechanisms can also affect it.

[0007] Position drift is manifested in the form of displacement, image distortion and / or distortion, which may significantly affect the predefined writing or image field of the particle beam. Therefore, it is usually necessary to correct or monitor the particle beam (i.e., drift correction / drift monitoring must be performed) to ensure that the particle beam is scanned along the desired working area.

[0008] The effect of the relative displacement between the mask and the particle beam is often addressed, for example, by defining the repair relative to a reference marking in the image or scanning area of ​​the particle beam, tracking the position of the reference marking over the duration of the repair by image processing methods, and displacing the particle beam for the repair by an offset. However, the marking or reference marking is itself subject to similar conditions as the defect to be repaired, and therefore degrades in the course of the repair process. Therefore, the marking is scanned as infrequently as possible, but as often as necessary.

[0009] US2002122992A discloses a method for correcting a photomask, in which a rectangular reference hole is formed in a mask pattern by an ion beam. By determining the position of the reference hole, it is possible to calculate the positional relationship between the reference hole and a defect in the mask pattern, and the defect is corrected using this positional relationship.

[0010] US20090218488 discloses a method and apparatus for beam correction of beam drift. The beam position is aligned with a marker close enough to the working area that it is possible to record an image of the marker without moving the positioning stage. The particle beam is corrected during processing by using a model that predicts the drift.

[0011] However, known techniques do not always perform satisfactorily, especially when repairing or inspecting large defects or defects having a large area (eg, greater than about 400×400 nm).

[0012] First, the efficiency of the repair process is reduced for such large area defects due to the absorption and diffusion processes of the process gas involved behaving differently. In the conventional approach, in addition, even for such large area defects, the process time increases with the increase in defect size, so that the fiducial marks need to be scanned more frequently, which results in them being further deteriorated. Such deteriorated fiducial marks cause more noise in the position determination, which results in the quality of the repair being impaired or even the repair being interrupted. In addition, the distance of the fiducial marks from the critical edge of the mask (or other relevant structures) is increased. As a result, the accuracy of edge positioning is reduced due to the increase in nonlinear effects caused by charging.

[0013] The present invention is therefore based on the object of specifying a method and an apparatus that provide improved possibilities for inspection and / or processing of lithographic objects (e.g. masks), in particular for inspection and / or processing of large area defects (e.g. having a diameter of more than a few hundred nm) on such objects. Summary of the Invention

[0014] This objective is at least partially achieved by various aspects of the present invention.

[0015] A first aspect of the invention relates to a method for generating correction marks on a lithographic object (e.g., a mask) using a particle beam, in particular for calibrating operations. The method can include the following steps: (a.) generating a first group of correction marks; and (b.) generating a second group of correction marks, where a separation of the correction marks within the first group and within the second group is smaller than a separation between a correction mark from the first group and a correction mark from the second group. In this case, the first group includes two or more correction marks, and the second group likewise includes two or more correction marks.

[0016] In this case, the separation between two correction marks results from the conventional scheme of determining the distance between two geometric objects. By way of example, this distance can be defined by the length of the shortest connecting line that spans between two points, each point belonging to a different correction mark. By way of example, the points in this case can include points on the edge of the correction mark (e.g., on the outer or inner edge of the correction mark) and / or can be any point that can be assigned to the area / geometry of the correction mark. For this reason, the correction marks in a group are strictly spaced from each other (e.g., placed next to each other) such that they are always closer to each other than the correction marks of the respective other groups.

[0017] In this case, the separation of the correction marks within a group can be referred to as the intra-group separation, whereas the separation between the correction marks from the first and second groups can be referred to as the extra-group separation. In this case, the intra-group separation includes all possible separations of the correction marks within a group. The extra-group separation includes all possible separations between any (desired) correction mark from the first group and any (desired) correction mark from the second group. Thus, no intra-group separation should be greater than the extra-group separation.

[0018] Known methods for generating correction marks on a mask have not previously relied on this type of geometric arrangement of the correction marks. Thus, operations on a mask that require correction marks for a specific purpose were previously only possible with multiple disadvantages. In this case, the operations may include, for example, (particle beam based) processing of the mask, inspecting the mask (e.g., with an electron beam of a scanning electron microscope, with an ion beam, etc.), locating the working area, calibrating the particle beam, etc.

[0019] In particular, an optimized arrangement of the correction marks can serve to determine the position drift of the particle beam in an improved manner and, as a result, to enable an optimized inspection and / or repair of errors on the mask (e.g., mask defects). Mask errors represent deviations of the mask from a corresponding target value and may arise, for example, during mask production. Common mask defects are sites or areas where there is too much or too little absorber material and are repaired, for example, by local etching with a particle beam or material deposition while the corresponding precursor gas is introduced.

[0020] In particular, the marks are thus provided locally (in the form of a first or second group), so that a calibration can be performed on the basis of the marks. Thus, when a mark of the first (second) group wears out, it is possible to resort instead to a further mark of the first (second) group, which is arranged nearby and can thus be used for further calibration of operations over a relatively long period of time (for example when processing and / or imaging relatively large defects). Alternatively, the marks of the group can also be used in a rotating manner, for example in order to distribute the wear over the marks and thus maintain all marks for as long as possible.

[0021] By way of example, the particle beam may be a beam of particles that have mass (e.g., an electron beam or an ion beam) or may otherwise be a massless particle beam (e.g., a photon beam).

[0022] The optimized operation enabled by the generation of groups according to the invention is explained in more detail in the second aspect of the invention.

[0023] In a further example, the first group and / or the second group include correction marks having at least partially the same form. For example, it is possible to generate correction marks of similar geometric types in the two groups (i.e. geometrically similar correction marks), which can then be used for the same function in the operation. This concept facilitates the presence of redundancy and / or dispersion of wear manifestations, for example when correction markers wear, fall off, and / or become no longer technically usable during the operation for technical reasons. In this case, the functions of the correction marks can include reference marks for the calibration of the operation, for example marks for determining drift of the particle beam, focusing marks for focusing the particle beam, localization marks for localizing the working area, alignment marks for aligning the operation (e.g. for aligning a mask for processing), verification marks for verifying the operation, etc. In this case, both the geometry and the material of the correction marks can be designed to be optimized for the specific function intended to be adopted by the correction marks. In this case, any material that can be (additionally) attached to the mask can be envisaged as material (e.g. metals such as platinum, tungsten, etc., and / or insulating materials), and the correction marks can likewise be defined from the mask material and / or the associated substrate (e.g. the correction marks can be formed from chromium, tantalum nitride, molybdenum silicide, the absorber material of the mask, etc.). Envisaged geometric shapes include, for example, circular, spherical, ovoid, triangular or rectangular correction markers, but also more complex geometric shapes such as crosses, frames, polygons, etc. As an example, a group of cross-shaped correction markers can be designed for a first localization of the working area, whereas the group of circular correction markers can serve for the calibration of the operation.

[0024] In a further example, the correction marks can be positioned next to each other. For example, the correction marks can be attached to the mask in rows and / or styles of an array (e.g., in a field having multiple adjacent rows of correction marks). In another example, only geometrically similar correction marks are generated in the first group and / or the second group, such that similar correction marks can be used for multiple functions during operation.

[0025] In a further example, at least one correction mark from the first group and / or the second group is composed of a plurality of geometric shapes. For example, the correction mark can be composed of a number of separate geometrically similar shapes (e.g. from identical circular / spherical points), for example generated very closely next to each other, thus forming the correction mark. By way of example, the generation can be carried out such that at a certain magnification, for example required for reading the correction mark by image processing, a uniform continuous correction mark recess is generated by the (for example very small) dimensioned separate shapes. The procedure according to the invention for generating a correction mark from a plurality of separate shapes can be advantageous from a process-technical point of view for the processing of lithographic masks. Firstly, an optimized process control occurs in this type of generation of correction marks, since the variability of the process conditions (and therefore the susceptibility to errors) can be minimized. This is due to the fact that the generation of the correction mark can be obtained by a number of similar (separate) individual processes, for example each individual process generating the same geometric shape on the mask and thus requiring the same process parameters to a large extent. Only the locations on the mask where the same geometric shapes should be applied need to be adapted for each step. This results in reduced process complexity, minimized process variations, and increased process stability, so that it is possible to generate correction marks that reliably have the same characteristics on the mask over a relatively long manufacturing period. This ensures that the correction marks function reliably for operations based on the correction marks, for example. Furthermore, the multiple distinct geometric shapes allow a simplification of the process of removing the correction marks, since smaller areas of the correction marks are connected to the mask. In addition, removal may be easier, since gaps with, for example, mask material are exposed as a result of the multiple distinct connection areas of the multiple geometric shapes to the mask in the correction marks. Thus, the removal process (e.g., cleaning by wet chemical processes, plasma processes, etc.) may act more uniformly on the correction marks, since, for example, edges of the geometric shapes in the correction marks may be eroded uniformly.This allows for minimal residue following the removal process.

[0026] In a further example, the correction mark can be composed of multiple more complex geometric shapes that are not necessarily geometrically similar. By way of example, rectangles in a frame, circular structures in a circular frame, reticles, etc. are conceivable in this case. This allows, for example, the creation of correction mark shapes required for a particular image processing method.

[0027] In an example, the number of correction marks from the first group and / or the second group is at least 3, preferably at least 4. This is particularly advantageous, for example for operations based on the correction marks, since it provides a suitable redundancy and / or a suitable wear distribution. In this case, the number of correction marks can be optimized for operations based thereon, such that as many correction marks as necessary, but as few as possible, are generated. This avoids unnecessary generation of correction marks, for example to minimize residues during removal.

[0028] In an example, the separation of the correction marks in the first group and / or in the second group is smaller than the separation between the correction marks from the first group and the correction marks from the second group, at least by a factor of 5, preferably by a factor of 10, particularly preferably by a factor of 20. As a result of the abovementioned difference of several orders of magnitude in the separation, for example, a quick group assignment of the two correction marks can be performed during the operation. The working area of ​​the operation based on the correction marks can be located, for example, in the area spanned by the separation between the correction marks from the first group and the correction marks from the second group (i.e. by the outer group separation). In this case, the characteristics of the correction marks can be varied depending on the chosen or generated dimensions of the outer group separation, for example, to optimize the readout of the correction marks. By way of example, if the separation of the correction marks in a group (i.e. the inner group separation) is at least a factor of 20 smaller than the outer group separation, the group can have a correspondingly larger area of ​​the correction marks. In contrast, if the inner group separation is a factor of 5 smaller than the outer group separation, the group can have, for example, a correction mark with a smaller area in comparison thereto. This can also be assumed for the type of correction mark (eg, its material, geometric shape, etc., as explained above) to be chosen and generated depending on the dimension of the outer group separation.

[0029] In an example, the generation of the correction mark is at least partially based on a particle beam induced deposition process and / or a particle beam induced etching process. In this context, a particle beam induced deposition process can include, for example, an electron beam induced and / or an ion beam induced deposition process. By way of example, the deposition process can be based on deposition (e.g., chemical and / or physical vapor deposition), which is induced or supported in a targeted manner by a particle beam. In this case, it is possible to locally deposit any desired geometric shape of the correction mark made of a suitable material on a mask by targeted guidance of, for example, a particle beam and a suitable gas atmosphere. Suitable methods known to those skilled in the art include, for example, (focused) electron beam induced deposition ((F)EBID), (focused) ion beam induced deposition ((F)IBID), etc.

[0030] Particle beam induced etching processes can include, for example, electron beam induced and / or ion beam induced etching processes. In this context, the etching process can take place, for example, in an etching environment (e.g., etching gas, etching plasma, etc.), and the etching process can be induced or assisted by a particle beam in a targeted manner. By way of example, any desired geometric shape of the correction mark can be locally etched into a mask material (e.g., absorber material of the mask, substrate) by targeted guidance of a particle beam in the etching environment. By way of example, (focused) electron beam induced etching ((F)EBIE) is a suitable method.

[0031] Furthermore, particle beam induced (e.g. ion beam induced) milling processes can also be used in examples, within which there is, for example, localized material ablation by induction of a particle beam. By way of example, ion beam milling is a suitable method.

[0032] In a further example, the method further comprises generating at least one third group of correction marks, the separation of the correction marks within each group being smaller than the separation between the correction marks from two different groups. Thus, the concept according to the invention is extended by at least one additional group. The generation of at least one fourth group of correction marks and / or at least a fifth group of correction marks, etc. are likewise conceivable in the examples.

[0033] In this case, the separation of the correction marks within a group can again be referred to as the intragroup separation, whereas the separation between correction marks from two different groups (e.g. the first and third groups, the first and fourth groups, the fourth and fifth groups, etc.) can be referred to as the extragroup separation. In this case, the intragroup separation includes all possible separations of the correction marks within a group. The extragroup separation includes all possible separations between (any) two correction marks from two different groups. The outline according to the invention is therefore based on the fact that the intragroup separation is not larger than the extragroup separation, even in the case of multiple groups (e.g. three and / or four groups of correction marks).

[0034] In particular, in an example, four groups can be generated. These can be generated at least partially in a quadrilateral arrangement (e.g. in the form of a square or rectangle). The quadrilateral can be arranged around the working area (e.g. the defect), i.e. the (geometric) centroids of the correction marks of the individual groups can form the corners of a quadrilateral that encloses the working area. In another example, the arrangement can be performed such that at least one rectangle can be formed, in which the working area is located, made, for example, from four correction marks from four different groups.

[0035] The working area can be considered as an area of ​​the object that is intended to be inspected or processed during operation. By way of example, this can be, for example, an area in which defective material is located. Within the scope of the present invention, the working area can in particular represent an area of ​​the object that is intended to be inspected / processed uniformly, i.e. an area that is neither inspected nor processed in multiple procedures that can be considered independent of each other. The working area of ​​a photomask is in particular a local area that does not encompass the entire photomask. For this reason, the working area can represent or include, for example, a local defect that does not, for example, decompose into a plurality of completely separate partial defects or that can be subdivided into completely separate partial defects. In this case, a "local" defect can mean, for example, small compared to the size of the object, for example the mask. For example, the working area or the defects contained therein can have one or more lateral extents of less than 1 mm, less than 100 μm, less than 10 μm, less than 2 μm, less than 1 μm or less than 500 nm. The working area or local defect may be fitted to a (virtual) square with side lengths of, for example, 1 mm, 100 μm, 10 μm or 2 μm or 1 μm or 500 nm.

[0036] In this case, the working area may be bounded by a peripheral edge or contour, and everything within this edge or contour may be considered the working area. In metaphorical terms, imagine, for example, a "lasso" that is cast around the defect and tightened until it "captures" the defect at a minimum length, where the "lasso" then represents the outer edge of the working area (alternatively, one may consider an elastic band that is cast around the defect and tightened around it, where the band, due to its width (e.g., on the order of the typical diameter of the fiducial markings discussed here), defines an edge strip around the working area rather than a linear edge). By way of example, the working area may be defined as an envelope of one or more areas to be repaired, e.g., a convex sleeve, and a group of compensation marks may be positioned such that they enclose the working area.

[0037] In a further example, the first and second groups of correction marks can be generated such that the envelope, e.g., convex sleeve, of at least one of the groups of correction marks does not intersect the working area (or such that the convex sleeve of at least one group of correction marks does not represent and / or form a subset of the working area (e.g., of a defect)). By way of example, the groups of correction marks can be positioned such that the convex sleeve of the first group and the convex sleeve of the second group do not intersect the working area (e.g., of a defect). This can be further interpreted as an area defined by the convex sleeve of the groups of correction marks does not form part of the working area (e.g., of a defect).

[0038] The rectangular arrangement can be designed as a calibration window for the calibration of an operation, for example for the compensation of distortions of a writing or image field processed with a particle beam, where the rectangular arrangement facilitates, for example, digital data processing or suitable image processing required for the compensation of distortions.

[0039] In further examples, the calibration window may be provided in the form of a triangle (i.e., with three generated groups) and / or with different types of polygons (i.e., as an m-polygon, where m is the number of groups generated).

[0040] In a further example, the generated groups surround the working area of ​​the operation in each case such that the connecting lines between two correction marks of different groups can surround the working area without intersecting the working area. As an example, the working area can be a mask error (e.g., a mask defect), and the connecting lines between two correction marks of different groups do not intersect the mask error. Furthermore, this can also be formulated as the connecting lines that define the outer group separation do not intersect the working area. This is advantageous, for example, since it ensures that the working area is completely surrounded by the groups (and the correction marks located therein). As an example, this allows a calibration window (e.g., as described above for distortion compensation) to completely encompass the mask error such that the entire working area can be interpolated over the calibration window without reducing the accuracy of the edge position. For example, measurement errors that may occur during the determination of the position of the correction marks (e.g., which may increase the extrapolation area) are consequently not amplified during distortion compensation.

[0041] In a further example of the method, the mask is first analyzed (e.g., by being scanned by a particle beam) to find and / or locate the working areas on the mask. By way of example, this may include searching for mask errors with the aid of image processing (e.g., by means of defect monitoring using pattern recognition). Information regarding the location of the working areas may then be used to mount the groups according to one of the examples described above.

[0042] In a further example, a method for generating correction marks on an object for lithography can include the following step: generating at least a local group of spaced apart correction marks. By way of example, it is possible to generate exactly one local group of spaced apart correction marks, i.e. it is not essential that two groups are generated. It is conceivable that the correction marks of the local group are generated in a local accumulation, the dimensions of the envelope of the correction mark (e.g. the extent of the correction mark envelope) being an order of magnitude smaller (e.g. shorter) than the dimensions of the envelope of the working area. By way of example, the dimensions of the envelope of the correction marks of the exactly one local group can be at least a quarter, preferably at least a tenth, most preferably at least a twentieth, of the dimensions of the envelope of the working area. In a further example, the convex sleeve of the local group of correction marks does not intersect the working area (or, for example, the convex sleeve is not a subset of the working area of ​​the operation). In this case, the separation of the correction marks in the local group can be designed in such a way that they are advantageous for the second aspect of the invention (more on this later). If only the generation of one local group is envisaged, this method can be combined with the further steps described herein, however, it is also possible to generate more than one local group independently of this.

[0043] A second aspect of the present invention relates to a method for calibrating operations on a lithographic object (e.g. a mask) using at least one local group of spaced apart correction marks using a particle beam, the method comprising the following sequences: (S1.) selecting a sequence tuple, the sequence tuple including a subset of the correction marks of the at least one group; (S2.) performing a calibration based at least in part on the sequence tuple; and (S3.) performing at least a portion of the operations based at least in part on the performed calibration.

[0044] By way of example, the second aspect of the invention may be based at least in part on at least one local group being a group generated according to any one of the above examples (i.e., a group generated according to the first aspect of the invention). In an example, the second aspect of the invention may include the first aspect of the invention (or, for example, the first aspect of the invention may also include the second aspect). By way of example, the method of the second aspect of the invention may include that the correction mark is generated (or has been generated) according to the first aspect of the invention. It is equally conceivable that (naturally) existing structures on the mask (e.g., one or more prominent points, structural transitions, various lithographic marks, etc.) may be used as the local group of correction marks.

[0045] At least one group is localized and thus arranged on the mask in a spatially delimited area. The spatial delimitation or extent can be provided, for example, by an envelope around the outer contour of the correction marks of the at least one localized group. In this case, the spatial delimitation can be related to the spatial extent of the operation (for example, the at least one localized group of correction marks can extend to an extent that is technically perceptible for the operation). Furthermore, the spatial extent of the localized group can be constrained, for example, to a continuous area, with the following area dimensions possible: 30 nm x 30 nm or less (or 9 x 10 -16 m 2 100nm×100nm or less (or 10 -14 m 2 or less), and / or 1mm x 1mm or less (or 10 -6 m 2 Moreover, the spatial extent of a local group can be, for example, the total area A M For example, the spatial extent of a local group can be defined as the ratio of its area to its area. -14 ×A M Below, 10 -12 ×A M Below, 10 -9 ×A M Less than or equal to 10-6 ×A M It may be: By way of example, the spatial extents mentioned above may also be applied to the generated groups described in the context of the first aspect of the invention. The local groups may also be distinguished in that the dimensions of the envelope of the correction marks (e.g. the perimeter of the envelope around the correction marks) are an order of magnitude smaller than the dimensions of the envelope of the working area. By way of example, the dimensions of the envelope of the correction marks of the local group may be at least a quarter, preferably at least a tenth, most preferably at least a twentieth of the dimensions of the envelope of the working area.

[0046] In this case, the sequence of the method according to the invention can be repeated several times within the scope of the overall operation. For this reason, calibration can be performed for different parts of the operation (i.e. sequence operations) in order to ensure stable parameters during the operation (i.e. within the sequence). The (overall) operation in this case includes the overall (i.e. self-contained) processing of the mask, whereas parts of the operation can be partial steps of the (overall) operation. In this case, the operation can include a particle beam-based procedure, for example the recording of an image field or the processing of a writing field with a particle beam can be envisaged. In this case, the recording of an image can include, for example, a recording by a scanning electron microscope, whereas the processing of a writing field can include a particle beam-induced etching or deposition process. In this case, the working area can include the writing or the image field. As an example, the calibration can be used to calibrate or correct the parameters of the particle beam for the operation.

[0047] To date, it has been known to deal with particle beam calibration, for example, by defining operations on identical correction marks. During calibration, the appropriate correction marks are typically scanned (e.g., with respect to the recorded image) with the particle beam. However, the corresponding correction marks and their immediate vicinity may be significantly degraded over time during an operation or during the entire operation, which may lead to an increasingly worse calibration. This is due to the fact that the correction marks being calibrated may be exposed to the same process effects of the operation (e.g., conditions of particle beam induced deposition or etching processes are still present). A disadvantage of previously known procedures is that a continuously reliable calibration is limited, in particular, by the processing duration of the operation. For longer processing operations (e.g., for relatively large working areas, working areas that can only be processed at low speeds, complex working areas, etc.), it is usually necessary to perform the calibration more frequently, which may cause the correction marks and their surroundings to deteriorate to a substantially greater extent. The accuracy with which the positions of the correction marks are determined by image processing may decrease significantly over time (e.g., as a result of reduced contrast, correction marks appearing washed out, etc.) In addition, permanent damage (e.g., defects) can remain on the mask following degradation behavior around the correction marks.

[0048] The concept according to the invention improves the known disadvantages due to the sequence tuples comprising a subset of correction marks from a group of correction marks selected for each sequence. This subset can then be used for calibration within the sequence. The calibration of the sequence is therefore based on a targeted selection of a subset of correction marks selected with a view to optimal calibration. In this case, the subset can comprise at least one correction mark (however, multiple correction marks from, for example, a group can also be envisaged). As a result, a selection of multiple correction marks (or one correction mark) suitable for the sequence will be possible. As a result, it is not necessary to resort to fixed correction marks for the calibration, which may for example already be significantly degraded, but it is possible to use a subset of correction marks that work well for the calibration. This ensures that the operation based on the calibration (part of the operation) can be performed optimally and / or without troublesome effects, since incorrect calibration of the operation is avoided.

[0049] In an example, the selection of the sequence tuple is based at least in part on the evaluation of a predefined criterion associated with at least one correction mark of at least one group. In this case, the predefined criterion can provide an indication of the degree of deterioration of the correction mark or can be designed such that the degree of deterioration is minimized. By way of example, the criterion can include the number of times the correction mark (or sequence tuple) has been used during the method, which can correspond to an estimate of the degree of wear of the mark. In particular, it is possible to evaluate how often the sequence tuple has been selected in succession. In a further example, the criterion can be evaluated by physical analysis (e.g. by recording an image of the correction mark using a particle beam).

[0050] In an example, the predetermined criteria include at least one of the following criteria: a wear degree of the at least one correction mark, a contrast of the at least one correction mark, a gradient image of the at least one correction mark, an autocorrelation function of an image of the at least one correction mark, a cross-correlation function of at least two images of the at least one correction mark.

[0051] By way of example, the above mentioned criteria may be determined based on a recorded image (e.g. a scanned electron image) of the correction mark (or the criteria may be determined based on a basic signal generated by scanning the correction mark with a particle beam). By way of example, the degree of wear may be measured or evaluated automatically by image processing (and / or manually by an operator). By way of example, a gradient image may be obtained by (digital) processing of the image (or of the basic signal) of the correction mark and may contain information about the edge of the correction mark. By way of example, this allows a good evaluation of the degradation of the edge, which may flatten out over time. Similarly, the characteristics of the autocorrelation function of the correction mark may be determined from the image (or from the basic signal) of the correction mark. By way of example, a wider autocorrelation function may in this case be associated with a higher degree of degradation, so that the corresponding correction mark may no longer be used for calibration above a given threshold. Furthermore, there may be a cross-correlation function of the correction mark based on the image (or basic signal) of the correction mark at two different times (e.g. from two different sequences). Thus, the relative changes in the correction mark may be evaluated analytically.

[0052] In a further example, the sequence is repeated at least once, and in the process at least two sequence tuples are selected that include different subsets of the correction marks. As a result, for the entire operation (e.g., including at least two sequences), the same correction marks do not have to be used consecutively for the calibration. This can significantly minimize both the manifestation of correction mark degradation and the mask damage caused by the calibration, since not all correction marks (or always the same correction marks) need to be exposed to the particle beam for every sequence. This ensures a reliable operation of the calibration, even with relatively long processing times that arise, for example, in the case of a large number of sequences.

[0053] In an example, a sequence tuple is selected to include a subset of correction marks from each of at least m groups of correction marks, where m is 2 or more (e.g., m=4), and the separation of correction marks within an individual group is less than the separation between correction marks from two different groups. By way of example, this example can be based at least in part on at least m groups, which are groups generated according to the first aspect of the invention (an example of which is described above). By way of example, four groups can be arranged in a rectangle on the mask, each group including, for example, at least three correction marks. This allows a particular sequence tuple to be chosen for calibration of the sequence from a large number of possible sequence tuples. By way of example, each tuple can include (exactly) one correction mark from each of the m groups.

[0054] In an example, the selection of sequence tuples is based at least in part on the expected number of sequences in the operation. By way of example, this can be based on the expected number of calibrations and / or portions of the operation. The number (of sequences, calibrations and / or portions of the operation) can be determined by, among other things, analyzing the working area. By way of example, this can include an analysis of mask defects, where several parameters of the mask defects, such as area of ​​the defect, type of defect, material of the defect, mask type, etc., can be observed to determine the number.

[0055] In examples, the selection of sequence tuples over the sequence of operations is performed according to a predefined order. The order can be defined such that the manifestations of wear (or the manifestations of deterioration) on the correction marks are minimized or uniformly distributed. In this regard, as a result of the skilled ordering of the sequence tuples, in examples, a physical analysis of the correction marks can be omitted (e.g., by evaluating appropriate criteria, as presented in some examples). Similarly, examples can be envisaged in which the sequence minimizes (and / or uniformly distributes) the manifestations of wear, but there is an accidental physical analysis of the correction marks (e.g., by recording an image in a subsequent analysis).

[0056] In examples, the selection (of sequence tuples) is performed according to a cyclic order of sequence tuples, a randomized order of sequence tuples, and / or rows of the same sequence tuple. As an example, multiple different sequence tuples (T1, T2, ... Tm) can be available on the mask. In the cyclic order, a certain cyclic order of sequence tuples is "rotated" across multiple sequences of methods. As an example, if three sequence tuples are available (e.g., T1, T2, T3), an exemplary cyclic order is given by T1-T2-T3, and the corresponding cyclic sequence order across the sequence of methods is, for example, T1-T2-T3-T1-T2-T3-T1-T2-T3-..., etc. In the case of three sequence tuples, the randomized order includes a random order of sequence tuples across multiple sequences of methods (e.g., T3-T1-T3-T3-T2-T1-T1-T2-T2-..., etc.).

[0057] When selecting sequence tuples according to the rows of the same sequence tuple, an exemplary sequence (in the case of three sequence tuples) can be determined in the process as follows: T1-T1-T1-...-T2-T2-T2-...-T3-T3-...-T3. In particular, when selecting sequence tuples according to the rows of the same sequence tuple, there may be an additional evaluation of a predetermined criterion (as described above in the exemplary form), which is also conceivable in the case of other orders. By way of example, the evaluation of the predetermined criterion can be performed after a certain number of sequences with the same sequence tuple to verify the state of the correction mark of the sequence tuple. Depending on the situation, depending on the state of the correction mark of the sequence tuple, it is possible to change to a different sequence tuple in the subsequent sequence (to ensure optimal calibration) or to continue using the sequence tuple (for example, due to the absence of wear-out manifestation yet).

[0058] In an example, the method further includes determining a transformation of particle beam parameters associated with the two sequence tuples and performing a calibration based at least in part on the determined transformation. In principle, at least two sequence tuples must be available for this concept, and thus the specified procedure can be performed for all possible sequence tuples (e.g., when more than two sequence tuples are available).

[0059] By way of example, the method may in this case include the determination of initial positions of the correction marks of the (possible) sequence tuples that can be selected for the method. In this case, the determination may be performed, for example, before a first sequence of the method. The mask may be electrostatically charged if the mask (e.g. the correction marks of the sequence tuples) is illuminated with a particle beam in the method. As a result, there are disturbances of the system that may affect the desired trajectory of the particle beam. Typically, the desired trajectory can be adjusted both dynamically and statically by particle beam parameters (e.g. acceleration voltages, deflection voltages, adjustments to particle beam optical elements, etc.). As a result of the above-mentioned disturbances, the positions of the (possible) sequence tuples may appear displaced relative to their initial positions during the course of the method (e.g. when further images are recorded by the particle beam). In this case, these "new" positions may be referred to as current positions. In this case, the offsets of the (possible) sequence tuples are not necessarily uniform and the correction marks may be or appear to be displaced independently of each other (e.g. due to different charging). By way of example, the correction marks of the (possible) sequence tuples may initially be positioned next to each other in a row such that their centroids form a straight line. However, during the course of the method, the correction marks may be displaced during the recording of the image such that their centroids no longer form a straight line (e.g., in a first perspective, each correction mark may appear to be randomly offset in space). These (e.g., non-linear) effects may affect the calibration of the particle beam such that a previously desired trajectory (e.g., over the working area) is no longer traversed. Thus, according to the invention, it is possible to perform a transformation that considers, for example, the initial position of one (or both) of the two sequence tuples relative to the current position of this and / or the other of the two sequence tuples. The transformation may thus re-establish the correct relationship between the two sequence tuples, which is no longer given diagrammatically as a result of a disturbance in the system.An in-sequence calibration can then take into account the information of the transformation so that the particle beam can be calibrated accordingly. This allows the particle beam trajectory to be independent of changes in the sequence tuples chosen in the method. This can ensure that the same working area is at least partially traversed by the particle beam. The transformation allows the quality of the edges of the working area in particular to be optimized, since the edges of the working area are not optimally processed in the case of missing transformations (e.g., flatter sidewall angles may result in etched and / or deposited structures).

[0060] In an example, the method further includes: determining a transformation of particle beam parameters associated with the selected sequence tuple and / or the selected sequence tuple in the preceding sequence, and / or performing a calibration based at least in part on the determined transformation. By way of example, the transformation may be performed for each sequence change. In another example, the transformation may be performed when changing a sequence tuple during the method (e.g., when changing a sequence tuple due to a chosen order, as described above).

[0061] In an example, determining the transformation includes at least one of the following: determining relative positions of the correction marks of the different sequence tuples with respect to each other and / or determining relative positions of the correction marks of the different sequence tuples with respect to one or more structures of the lithographic object (e.g., mask). In this case, the relative positions (or position differences) may be determined, for example, in accordance with a style of vector (e.g.,<x,y> )

[0062] As an example, the relative positions between the initial positions of the correction marks of the (possible) sequence tuples can be determined (for example, in each case between two correction marks of sequence tuples from different groups). Furthermore, the relative positions between the current positions of the correction marks of the (possible) sequence tuples can also be determined. A transformation can then be determined from the different relative positions that allows the change from one sequence tuple to the other and minimizes the above-mentioned errors in the process. Furthermore, it is also possible to determine the relative positions between the initial positions of a sequence tuple and the current positions of different sequence tuples. Similarly, it is also possible to determine the relative positions between the initial position of a sequence tuple and the current position of this sequence tuple.

[0063] Furthermore, the above-mentioned determination of the relative positions of the correction marks of the various sequence tuples with respect to one or more structures considers at least one third "fixed point" that includes one or more structures (e.g., of a mask). Thus, a transformation can be identified by determining the position of the (characteristic) structure with respect to the position of the first sequence tuple and determining the position of the (characteristic) structure with respect to the position of the second sequence tuple. Thus, for example, there can be a transformation from a first sequence tuple to a second sequence tuple via a third "fixed point".

[0064] In examples, the calibration includes determining a drift of the particle beam and / or correcting the drift of the particle beam. In this case, the calibration can include determining an offset associated with the determined drift of the particle beam. The offset can then be used to correct the drift of the particle beam such that a desired trajectory of the particle beam is constant or maintained during the method.

[0065] In an example, the action includes repairing the defect. In this case, the defect may include, for example, a mask defect, and the defect may be treated such that the defect is at least partially repaired by a deposition or etching process (as described in the example above). Here, the working area may include the defect or may be defined by its dimensions (e.g., by its outline).

[0066] It is emphasized that the first and second aspects described herein can also be combined with each other, i.e. the generation of the correction marks described herein can be followed by the calibration method described herein, supported by the generated correction marks.

[0067] A third aspect of the present invention relates to an apparatus for generating correction marks on a mask and / or calibrating operations using a particle beam, the apparatus comprising (a.) means for (automatically) executing one of the methods described herein and (b.) means for executing a computer program. In this case, the apparatus may include an apparatus suitable for performing a particle beam-based process. In this case, the apparatus may include, for example, an electron beam system and / or an ion beam system configured for a particle beam induced deposition and / or etching process. Processing is also conceivable in a particle beam system capable of controlling at least one electron beam and also at least one ion beam in a targeted manner (e.g., a dual beam system or a cross beam system).

[0068] A fourth aspect of the invention relates to a computer program comprising instructions which, when executed, cause an apparatus as described above to carry out the method steps according to one of the methods described herein.

[0069] Further aspects relate to the above-mentioned device having a memory containing a computer program. Furthermore, the device can have means for executing the computer program. Alternatively, the computer program can be stored elsewhere (e.g., in the cloud) and the device can only have means for receiving instructions resulting from the execution of the program elsewhere. In any case, this can allow the method to be performed in an automated or autonomous manner in the device. As a result, it is possible to minimize, for example, the intervention of an operator, thus minimizing both the cost and the complexity when processing the mask. Furthermore, the method can also be present in written form (e.g., in an instruction manual for the above-mentioned device) to communicate, for example, targeted instructions regarding the method steps to an operator.

[0070] The following detailed description describes technical background and exemplary embodiments of the present invention with reference to the drawings. [Brief description of the drawings]

[0071] [Figure 1] FIG. 1 illustrates an embodiment of a problem that arises when using an electron beam to inspect and / or process photolithographic objects, where the elements have charged surfaces. [Diagram 2] FIG. 1A illustrates a schematic top view of an exemplary repair situation of a defect in a photolithography mask as described in the prior art. [Diagram 3] 1 is a graph that illustrates a schematic representation of compensation for drift of an electron beam relative to a marking caused by electrostatic charging, according to the prior art; [Figure 4] 3 is a graph reproducing the displacement of the markings relative to the x-axis and y-axis during the repair of the defect of FIG. 2. [Figure 5a] 5A and 5B show schematic diagrams of various aspects of the invention, in this case an example of a correction mark according to the invention, for calibrating the movement relative to a mask; [Figure 5b]5A-5C show schematic diagrams of various aspects of the present invention: Fig. 5b illustrates the selection of a sequence tuple for operation; [Figure 5c] 5A-5C are schematic diagrams illustrating various aspects of the present invention, and FIG. 5C illustrates the topic of transformation of particle beam parameters in the present invention. [Figure 6] Finally, it shows a schematic diagram of some components of an apparatus for carrying out the method according to the invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0072] Some technical background information and possible embodiments of the method and apparatus according to the invention are described in more detail below on the basis of the inspection of photolithographic masks and the treatment of defects in photolithographic masks.

[0073] The methods and devices according to the invention can be used for the first time to inspect and / or process all types of transmissive and reflective photomasks. Furthermore, the methods and devices according to the invention can also be used to inspect and / or process templates and / or wafers for nanoimprint lithography. The devices and methods according to the invention are furthermore not even limited in principle to inspecting and / or processing (photo)lithographic objects. Rather, they can be used in general to analyze and / or process non-conductive or only slightly conductive samples using a charged particle beam.

[0074] However, for the sake of clarity and to avoid ambiguity, the following embodiments relate throughout to the example of a photolithographic mask, however, other possible uses of the aspects of the invention described are always encompassed and should therefore also always be taken into consideration.

[0075] The diagram 100 of Fig. 1 shows a schematic cross-sectional view of a charged mask 110 and the output 165 of a scanning electron microscope 160. The mask 110 has on its surface 120 a distribution of surface charges which causes a potential distribution or electrostatic charging of the mask 110. In the image portion 105 on the left, the mask surface 120 has a positive charge 140. In the image portion 195 on the right, the mask surface 120 shows an excess of negative charges 150. Hereafter, the references 140 and 150 are used to indicate both the distribution of surface charges on the mask surface 120 and the potential distribution caused by the charged surface.

[0076] The charging 140, 150 of the mask surface 120 may be caused by a beam 170 of charged particles, e.g., the electron beam 170 of a scanning electron microscope (SEM) 160. The electrostatic charging 140, 150 of the mask surface 120 may be caused by scanning the mask 110 as part of an inspection process or may result from a processing process. For example, electrostatic charging may be caused during processing of the mask 110 with an electron beam or an ion beam. Additionally, the electrostatic charging 140, 150 of the mask 110 may be caused by, for example, handling of the mask 110.

[0077] In the portion of the mask 110 depicted in the schematic diagram 100 of Figure 1, the distribution of surface charges 140, 150 has a uniform density, however, this is not a necessary condition for the explanations given herein.

[0078] 1, a deflection system 175 deflects and scans the electron beam 170 over the mask surface 120 to determine the dimensions of the structural elements 130 of the mask 110. By way of example, the structural elements 130 can be pattern elements of absorber structures of the mask.

[0079] As shown in the left image portion 105 of diagram 100, as a result of the attractive effect of the positive charge 140 on the mask surface 120, the electron beam 170 scanning the structural element 130 is deflected towards an optical axis 172 in the vicinity of the mask surface 120 and follows a trajectory 174. In the absence of the electric potential distribution 140, the electron beam 170 would follow a path 176. In the SEM image produced by the electron beam 170, the scanned dimension 178 appears larger than the actual dimension 180 of the structural element 130.

[0080] 1 shows the repelling effect of the negatively charged (150) mask surface 120 on the path travel 184 of the electrons 170 of the electron beam 170. Without the potential distribution 150, the electron beam 170 would follow path 186. As a result of the electrostatic charge 150, and as a result of the further deflection of the electron beam 170 directed away from the beam axis 172 in the vicinity of the mask surface 120, the measured dimension 188 of the structure element 130 in the SEM image generated from the scan data appears to have a smaller dimension than the actual dimension 180 of the structure element 130.

[0081] Scanning the structural elements 130 by the electron beam 170, or more generally with the charged particle beam 170, can result in local heating of the mask 110 and thus changes in the extent of the mask 110. Even if these changes in the length of the mask 110 are only on the order of nanometers, these changes should be taken into account in the treatment process of the mask 110 in order not to jeopardize the success of the treatment process. Furthermore, thermal effects of the SEM 160 and / or the mask 110, or the sample mount (not shown in FIG. 1), can again cause the point of incidence of the electron beam 170 on the mask 110 to drift in the double-order nanometer range as a function of time.

[0082] Fig. 2 shows a part of a top view of a mask 200. This can be for example the mask 110 in Fig. 1. The photomask 200 comprises a substrate 210. Two pattern elements 220 and 230 in the form of absorbent strips are arranged on the substrate 210 of the mask 200. In the pattern element 220, the mask 200 has a defect 250 in the form of an excess of material. In order to correct the defect 250, in the example shown in Fig. 2, a marking 240 is applied to the pattern element 220. The marking 240 (which can also be called a correction mark) is used to determine and compensate for a drift or displacement of the electron beam 170 with respect to the defect 250 during the repair process of the defect 250.

[0083] The marking 240 is deposited after identification of the defect 250 on the mask 200, for example by means of an electron beam induced deposition (EBID) process, i.e. by supplying at least one precursor or process gas onto the mask 200. This is advantageous if the precursor gas is selected such that the marking 240 has a different material composition than the pattern elements 220, 230 of the mask 200. In the SEM 160 image, the marking 240 further distinguishes itself by material contrast in addition to the topological contrast.

[0084] To remove the defect 250, an etching reaction is triggered due to the electron beam or particle beam at the location of the defect 250, for example by supplying an additional precursor gas or process gas (or gas mixture), which results in the removal of the defect 250. In line with this, and in line with what has been stated above regarding the meaning of the term "working area", the working area 260 in FIG. 2 is substantially defined by the extent of the defect 250 and is bounded or surrounded by a contour line 265. The working area is shown here in FIG. 2 only in a highly schematic form. However, as can be clearly seen, in the case of FIG. 2 showing the prior art, the marking 240 is located outside the working area 260.

[0085] For example, material deposition to correct specific imperfections in mask 200 is possible as well.

[0086] FIG. 3 shows, as an example, a schematic compensation of the drift or displacement of the electron beam 170 relative to the marking 240 during the repair process of the defect 250 according to the prior art. The local electrostatic charging of the mask 200 is difficult to define mathematically. This also applies to the thermal drift between the electron beam 170 and the marking 240. Therefore, the effect of the electrostatic charging of the mask 200 and / or its displacement relative to the point of incidence of the electron beam 170 relative to the marking 240 is measured and compensated for at periodic time intervals. The solid curve 310 in FIG. 3 shows a schematic change, displacement, variation or drift of the marking 240 as a function of time during the repair process of the defect 250.

[0087] At the start of the repair process, a reference position 330 of the marking 240 is determined. The reference position 330 can be specified relative to a reference marking of the mask 200 or absolutely relative to the coordinate system of the mask 220. In a second step, the position of the repair shape is defined relative to the marking 240. In this case, the repair shape can be designed to cover the defect 250 in the working area 260. In an example, the repair shape can correspond (at least in part) to the spatial dimensions of the defect 250 (e.g., the area, shape and / or contour of the repair shape can correspond to corresponding characteristics of the defect 250). In a further example, the repair shape can correspond to the working area 260 (e.g., the repair shape can be identical to the working area 260). Inspection or treatment of the defect 250 can be performed, for example, such that an electron beam is scanned along the repair shape, resulting in or causing inspection or treatment of the working area or the defect in a manner known per se. The repair shape may be known beforehand, for example, from inspection or processing of similar work areas or defects (eg, work areas or defects having generally the same size, generally the same shape, material properties, defect class, etc.).

[0088] Thereafter, the repair of the defect 250 is started. For this purpose, as already mentioned, one or more etching gases are supplied to the location of the defect 250 in Fig. 2 and an electron beam 170 is scanned over the defect 250 through a working area 260, shown diagrammatically in Fig. 2, as predefined by the repair shape.

[0089] After the lapse of a particular time interval 320, the repair process is interrupted at regular or irregular time intervals 340, but without interruption of the supply of precursor gas, in order to scan the marking 240 with the electron beam 170. A displacement, drift or change 350 in the marking relative to a reference position 330 or relative to a previous measurement of the marking 240 is determined from an SEM image of the marking 240. The position of the repair feature in relative or absolute terms to the marking 240 is then corrected based on the change 350 in the marking, and the repair process of the defect 250 continues.

[0090] FIG. 4 represents a further example of the displacement or drift of the marking 240 during the repair process of the defect 250 according to the prior art. On the x-axis of the graph 400 of FIG. 4, the time is plotted in arbitrary units. On the horizontal axis of the graph 400, the number of measurements of the marking 240 during the repair process can also be represented. The time interval between performing two scanning processes can be in the range of 1 second to 50 seconds. The example shown in FIG. 4 shows a time range of about 1000 seconds. On the y-axis of the graph 400, the total displacement or drift of the marking 240 is represented in arbitrary units relative to the reference position 330 of the marking 240. By way of example, the drift can be specified as the number of scanned pixels of the electron beam 170 in one direction. Depending on the focus of the electron beam, the pixels can have dimensions in the range of 0.1 nm to 10 nm. The ordinate of the graph 400 comprises a position change of about 120 nm.

[0091] The drift of the marking 240 in the x-direction is represented by curve 410 of the graph 400, and the displacement of the marking 240 in the y-direction is represented by curve 420. A large position change or displacement of the marking 240 is caused by switching between two process or precursor gases. This is indicated by arrow 440 in Figure 4. A smaller oscillation or jump in the position change is caused by, for example, switching between different repair shapes to repair the defect 250 (see arrow 430).

[0092] 2-4 may be adequate for small defects such as defect 250 (it should be noted that the extent of defect 250 is smaller than the dimensions of lines 220, 230 and the space between them, i.e., typically a few nm). However, for larger area defects (e.g., in the range of hundreds of nm), the situation changes drastically.

[0093] 5a to 5c show various aspects of the invention in a schematic way. Proceeding from the situation known from the prior art, the invention makes it possible, inter alia, to increase the accuracy and efficiency of the inspection and / or processing of masks, especially in the case of large area continuous defects.

[0094] In this case, FIG. 5a shows an example of a correction mark according to the invention for calibrating an operation on a mask and an associated working area (which can also be called marking, reference marking, etc.). In this case, the working area 500 can include a defect on the mask to be repaired. The operation can serve to repair the defect on the mask, and a particle beam-based process can be envisaged here (e.g. a particle beam induced deposition or etching process). In this context, the working area 500 can be located at the beginning of an initial step. For this, it is possible to use, for example, conventional methods for defect recognition on a mask. By way of example, the mask can be scanned with an electron beam so that a scanning electron image is generated. Image processing can then analyze the scanning electron image in order to locate the defect and in addition define the working area 500. By way of example, the image processing can be part of a method for automatic defect control and can include pattern recognition of associated defects. The working area 500 can be considered as a writing field (i.e. for physical processing of the mask) and / or an image field (i.e. as an object of recording an image of the working area).

[0095] According to the invention, at least two groups of correction marks can be generated around the working area 500. By way of example, the groups can be generated by a particle beam-based process. In this case, particle beam-induced deposition or etching processes can be envisaged, such as electron beam-induced deposition, ion beam-induced deposition, electron beam-induced etching, ion beam-induced milling, etc. The deposition or etching can be carried out here on the mask material or on the substrate of the mask. The application of particle beam-based processes can allow a large degree of freedom when designing the correction marks with regard to their geometry, their material and their position on the mask. Thus, the groups of correction marks around the working area 500 can be designed with various variants. By way of example, the correction marks can have any desired geometry, such as circular structures, an amalgam of closely packed circular structures, polygons, spheres, holes, grooves, etc. In this case, the geometry of the correction marks can crucially depend on the chosen deposition or etching process. By way of example, the further topology can be generated by a deposition process, whereas the structures can be generated by material ablation in an etching procedure. Similarly, the correction mark material may depend on the deposition or etching process selected. By way of example, the material of the correction mark may include metals (e.g. platinum, tungsten, silver, gold, etc.), insulators (e.g. nitrides, oxides, polyimides, etc.), semiconductors, etc. in a deposition process. Composite materials are also conceivable here. In the case of an etching process, the correction mark material may include or be defined by the material of the mask and / or the substrate. By way of example, the correction mark may be defined as an absorber strip of the mask, a hole (etched / milled) in the substrate of the mask, etc.

[0096] As presented in Fig. 5a, for example, four groups of correction marks can be arranged around the working area. In this case, the first group can include correction marks A1, B1, C1. The second group can include correction marks A2, B2, C2. The third group can include correction marks A3, B3, C3. The fourth group can include correction marks A4, B4, C4. The groups can each include the same number of correction marks (in this example, each group has three correction marks), and it is also conceivable that the groups can include different numbers of correction marks.

[0097] According to the invention, a group is defined by a local accumulation of correction marks spatially separated from different groups of correction marks. In this case, the separation of correction marks within an individual group (i.e., intragroup separation) can be smaller than the separation between correction marks from two different groups (i.e., extragroup separation). By way of example, the following separations between correction marks occur in the first group: A1B1 (correction marks A1 to B1), A1C1 (correction marks A1 to C1), B1C1 (correction marks B1 to C1). In this case, the separations can be determined, for example, by the centroids, edges, any surface points, etc. of the correction marks. The separations between correction marks from two different groups are, for example, A1A3 (correction marks A1 to A3), C1A2 (correction marks C1 to A2), A1C4 (correction marks A1 to C4). As is clear from FIG. 5a, the separation between correction marks from two different groups is smaller than the separation (any separation) of correction marks within a single group (eg, A1A3>A1C1).

[0098] The groups arranged in this way can serve to optimize the operations for the processing of the mask or the processing of the working areas of the mask. As explained in relation to FIG. 3, the calibration with the markings (or correction marks) must usually be performed several times for the operations during which the mask is processed. By way of example, this can be the compensation of drifts or displacements of the particle beam, and the repair shape can be adapted accordingly. During the calibration, the markings (or correction marks) must usually be measured with the particle beam in order to determine the position of said markings, as a result of which the markings (or correction marks) may deteriorate and / or wear (as explained above).

[0099] According to the invention, as previously known, not all markings (or correction marks) are necessarily provided for each calibration step of the (whole) processing of the mask (or of the corresponding working area of ​​the mask). The generated arrangement of the groups of correction marks allows targeted selection of the desired correction marks for the calibration (for example, in case of calibration during repair of defects in the working area of ​​the mask, this can be calibrated by markings, for example). In this case, the boundary condition for the calibration can be that at least one correction mark from each group is used. Thus, for the purpose of the calibration, it is possible to select from a set of correction marks so that the calibration step is not limited to a single sequence tuple of correction marks. According to FIG. 5a, a large number of possible sequence tuples arise with the specified boundary conditions (four sequence tuples are listed in an exemplary form: A1, A2, A3, A4; A1, B2, C3, C4; B1, A2, C3, A4; B1, A2, A3, A4; ...). Thus, in more general terms, it is possible to choose from a large number of different tuples T1, T2, . . . Tm for calibration.

[0100] According to the invention, the selection of sequence tuples allows minimizing or distributing the degradation of the correction marks during the process. By way of example, different sequence tuples can be used for the calibration steps during the course of the process. In the process, the degradation can be minimized according to several principles.

[0101] First, a statistical distribution of the use of sequence tuples can be envisaged. As an example, a fixed number of sequence tuples (e.g., three tuples: T1, T2, T3) can be made available. To ensure a uniform distribution, sequence tuples can be used periodically over the calibration steps in one variant. For example, it can be envisaged that the order of sequence tuples is reused periodically. In the case of three sequence tuples, the selection of sequence tuples over several calibration steps can now be seen as follows: T1-T2-T3-T1-T2-T3-T1-T2-T3-.... As a result, each sequence tuple is stressed to approximately the same extent over the process. Furthermore, it can also be envisaged that a randomized order of sequence tuples is selected. In this context, the type of order can be based on a mathematically uniform distribution, such that the sequence tuples used are used to approximately the same extent after a number of calibration steps.

[0102] Furthermore, it can be assumed that sequence tuples are always used consecutively until there is a change to a different sequence tuple. First, it can also be designed so that there is a statistically uniform distribution of sequence tuple usage. In this case, the number of predicted measurements (or calibration steps) during the processing of the mask can be estimated. As an example, it can be estimated that nine calibration steps are required for processing in one example. Thus, for example, in the case of three sequence tuples, the following order can be chosen in the calibration steps to ensure a uniform distribution of sequence tuple usage: T1-T1-T1-T2-T2-T2-T3-T3-T3.

[0103] Secondly, it can be envisaged that sequence tuples are used successively until a significant degree of wear occurs on the correction marks of the sequence tuples. When a significant degree of wear occurs, there can be a change to a different sequence tuple in the next calibration step. By way of example, a significant degree of wear can be ascertained by evaluation of a quality criterion of the correction marks of the sequence tuples (for example, a scanning electronic image recording must be performed for this purpose). In an example, all correction marks of the sequence tuples can be evaluated against a quality criterion. By way of example, the quality criterion can be, for example, the contrast of the correction marks. By way of example, this can also be the contrast difference between the correction mark and its immediate surroundings. In this case, in the case of an excessively strong or excessively weak contrast, a trigger can be prompted to initiate a change of the sequence tuple in the next calibration step. Furthermore, the quality criterion can be an evaluation of the gradient image of the correction marks. By way of example, if it is inferred from the gradient image that the edges of the correction marks are significantly degraded, it is possible to prompt a trigger to initiate a change of the sequence tuple in the next calibration step. Furthermore, it is also possible to analyse the properties of the autocorrelation function of the image of the correction marks. In this case, the threshold for prompting a trigger of a sequence tuple change can be the width of the autocorrelation function, which can also provide information about, for example, the state of the edge of the correction mark. Furthermore, it is also possible to analyze the properties of the cross-correlation function of the images of the correction mark at two different times. As a result, the correspondence of the images can be analytically evaluated and a measure of similarity of the images of the correction mark (at the various times) can be determined. As an example, if the measure of similarity of the images (at the various times) exceeds a threshold, it is possible to prompt a trigger resulting in a sequence tuple change. Furthermore, it is also possible to analyze the properties of the cross-correlation function of the images of different correction marks. As an example, there can be a cross-correlation between a correction mark of a sequence tuple and a correction mark of the same sequence tuple.Furthermore, there may be a cross-correlation between a correction mark of a sequence tuple and a correction mark not included in the (current) sequence tuple (e.g., this latter correction mark may be from a sequence tuple already used for calibration or from a sequence tuple not yet used for calibration). A combination of the autocorrelation and cross-correlation functions of one or more correction marks may also be used for the evaluation of the quality criterion. Furthermore, the quality criterion of the correction mark or the degree of wear of the correction mark may be estimated. By way of example, this estimate may be based on the number of calibration steps in which the correction mark has already been used during processing. By way of example, based on the utilized process (e.g., an electron beam induced deposition process), it may be estimated that, for example, after the correction mark has been called 10 times (i.e., for 10 calibration steps), a significant degradation of the correction mark may be expected, and therefore a change to a different sequence tuple should be made. By way of example, this estimate may be based on empirical values ​​and / or experiments.

[0104] FIG. 5b shows the selection of sequence tuples for the operation. In particular, three sequence tuples P1, P2, P3 are shown. With reference to FIG. 5a, in this example P1 is given by correction marks A1, A2, A3, A4 (references of the correction marks are not shown in FIG. 5b). In this case P2 is given by correction marks A1, C2, A3, B4. In this case P3 is given by A1, A2, A3, A4. The connecting lines between the correction marks of the groups are presented using dashed lines for each sequence tuple, while the connecting lines between the correction marks from the second group and the correction marks from the fourth group are labelled in more detail (501, 502, 503). According to the invention, the correction marks can be arranged such that the working area is located within a polygon (for example a triangle, trapezoid, rectangle, pentagon, etc.) spanned by the sequence tuple. By way of example, the correction marks can be arranged such that their convex sleeves define (or contain) the working area. In this case, this can be done so that the working area is completely surrounded by the connection lines and the connection lines of the sequence tuples do not cross the working area. As a result, this is the case where the calibration window (spanned by the correction marks of the sequence tuples) completely surrounds the envelope or contour of the working area 500. Thus, the whole working area, or the distortion of the particle beam relative to the working area, can be interpolated over the calibration window. By way of example, the calibration window can be used for distortion compensation, which can measure and compensate for the displacement of the particle beam as a first order aberration as well. By way of example, these situations are given for sequence tuples P1 and P2 of FIG. 5b, but not for sequence tuple P3. In sequence tuple P3, the connection line 503 from correction mark A4 to correction mark A2 crosses the working area 500, so that the crossing 533 is located outside the calibration window. By way of example, the distortion compensation (performed by the calibration window) performs an extrapolation in this case, which may reduce the accuracy of the edge position (e.g. of a defect). During processing of a work area (e.g., a defect), this may lead to the edges (or contours) of the work area (e.g., a defect) not being optimally traversed by the electron beam (e.g., not following a desired trajectory).These situations may further amplify measurement errors when determining the mask position. Thus, the generation of correction marks in the style of sequence tuple P3 can be avoided. If such sequence tuples are nevertheless generated (e.g. due to manufacturing variations), this type of sequence tuple can preferably be excluded by the calibration method or excluded from the formation of the calibration window. Alternatively, marks from one group can be selected based on marks chosen from another group, such that this type of sequence tuple is excluded.

[0105] FIG. 5c illustrates the topic of particle beam parameter conversion in the present invention. Four groups of the above-mentioned correction marks, for example applied to a mask, are shown (in a manner similar to FIG. 5a and FIG. 5b). The mask can be electrostatically charged during irradiation with a particle beam (for example, an electron beam). These charging phenomena have, among other things, a nonlinear effect on the particle beam, so that in the absence of technical adaptations, there is an impact on the trajectory of the particle beam when changing sequence tuples during the process. As an example, FIG. 5c represents the corresponding trajectories associated with sequence tuples. Without further intervention, these particle beam trajectories will be traversed during the calibration. In this case, trajectory 510 is associated with sequence tuples C1, C2, C3, C4, and trajectory 520 is associated with sequence tuples B1, B2, B3, B4. Furthermore, trajectory 530 is associated with sequence tuples A1, A2, A3, A4.

[0106] These effects do not necessarily lead to significant offsets of the particle beam (as shown diagrammatically in FIG. 5c). However, as diagrammatically represented by trajectories 510, 520, 530, the quality of the edges of the etched or deposited structures in the operation may be degraded due to the edges of the working area 500 not being identically traversed by the particle beam. As an example, this can be made perceptible by flatter sidewall angles of the structures. This phenomenon can be addressed by determining a mathematical transformation of each sequence tuple, which compensates for the differences between the sequence tuples. This transformation can be further taken into account during calibration (e.g., in addition to the distortion compensation of the repair shape). As a result, it is possible to ensure that the trajectory of the particle beam remains the same even in the case of calibration with different sequence tuples. First, the global position of all correction marks of the sequence tuples and / or the relative (initial) positions of all correction marks with respect to each other can be determined for the establishment of the transformation. In addition, it is possible to determine the corresponding positions of the characteristic mask structures. The relative position may be expressed in terms of vectors, e.g., in terms of x and y distances (e.g.,<x,y> The distance vector can be specified.

[0107] The transformation can then be determined in a number of ways, for example by determining the relative positions of sequence tuple X and sequence tuple Y. As an example, sequence tuple Y can initially be known to have a vector separation from sequence tuple X of V1=<50 nm, 0 nm> (e.g. defined by the separation between the (two or more) corresponding marks of the tuples). Initially, only sequence tuple X can be used within the scope of the invention. However, when changing from sequence tuple X to sequence tuple Y during the course of the method, it can now be known that sequence tuple Y has a vector separation from sequence tuple X of V2=<100 nm, 10 nm>. This information can be used for the transformation such that the particle beam can adapt accordingly to provide a constant trajectory (to compensate for the apparently changed separation).

[0108] As an example, the transformation can also be established by establishing the position of the (characteristic) mask structure relative to the sequence tuple X and the sequence tuple Y. Initially, only the sequence tuple X can be used within the method. Initially, the separation of the (characteristic) mask structure from the sequence tuple X can be given by V3=<500 nm, 500 nm> and from the sequence tuple Y by V4=<550 nm, 500 nm>. However, when changing from the sequence tuple X to the sequence tuple Y during the course of the method, it can now be known, for example, that V3=<600 nm, 400 nm> and V4=<700 nm, 350 nm>. This information can be used for the transformation such that the particle beam can be adapted accordingly to provide a certain trajectory.

[0109] Fig. 6 shows, in a cross-sectional view, some components of an apparatus 600, in which an embodiment of the method according to the invention for inspecting and / or processing a mask (or generally one of the objects mentioned in the introduction for which the invention can be used) can be carried out and implemented. By way of example, reference is made to Fig. 6 and the following description is given to the mask 510 of Figs. 5a-5c, which should not be understood as limiting. Instead, other lithographic masks or objects can be used.

[0110] The apparatus 600 comprises a vacuum chamber 602 and inside a scanning particle microscope 620. In the example of Fig. 6, the scanning particle microscope 620 is a scanning electron microscope (SEM) 620. An electron beam as particle beam has the advantage that the mask 510 to be inspected or processed cannot be substantially damaged by said beam or can only be damaged to a small extent. However, other charged particle beams are also possible, for example an ion beam of a FIB (Focused Ion Beam) system (not shown in Fig. 6).

[0111] The SEM 620 comprises as essential components a particle gun 622 and a column 624 in which an electron optical or beam optical unit 626 is arranged. The electron gun 622 generates an electron beam 628, which the electron or beam optical unit 626 focuses and directs at the output of the column 624 onto the mask 510 (or generally onto a lithographic sample or object). The mask 510 has a surface 520 with one or more structures 530, as already described in detail above. Surface charges that may be present on the mask 510 are not shown in FIG. 6.

[0112] The mask 510 is disposed on a sample stage 605. As symbolized by the arrows in Figure 6, the sample stage 605 can move in three spatial directions relative to the electron beam 628 of the SEM 620.

[0113] The spectrometer-detector combination 640 distinguishes the secondary electrons generated by the electron beam 628 at the measurement point 635 and / or the electrons backscattered by the mask 510 based on their energy and then converts them into an electrical measurement signal. The measurement signal is then passed to an evaluation unit 676 of the computer system 670.

[0114] To separate the energies, the spectrometer-detector combination 640 can include a filter or filter system to differentiate electrons by energy (not shown in FIG. 6).

[0115] Similar to spectrometer-detector combination 640, the energy-resolving spectrometer can be located outside of column 624 of SEM 620. However, it is also possible to locate the spectrometer and associated detector within column 624 of SEM 620. In the example shown in Figure 6, spectrometer 645 and detector 650 are integrated within column 624 of SEM 620. In addition to or as an alternative to spectrometer-detector combination 640, spectrometer 645 and detector 650 can be used in apparatus 600.

[0116] 6 may optionally include a detector 655 for detecting photons generated by the incident electron beam 628 at measurement point 635. The detector 655 may, for example, spectrally resolve the energy spectrum of the generated photons, thereby allowing conclusions to be drawn regarding the composition of a layer at or near the surface 520 of the mask 510.

[0117] In addition, the apparatus 600 may comprise an ion source (not shown) for providing low energy ions in the region of the measurement point 635, in case the mask 510 or its surface 520 is electrically insulating or semiconductive and has a negative surface charge. The ion source may be used to reduce the negative charge of the mask surface 520 in a localized and controlled manner.

[0118] If the mask surface 520 has an undesirable distribution of positive surface charge, for example caused by handling of the mask 510, the electron beam 628 can be used to reduce the charging of the mask surface 520.

[0119] The computer system 670 comprises a scanning unit 672 for scanning the electron beam 628 on the mask 510, in particular over the markings 540, 580 and / or the defects 550. The scanning unit 672 controls deflection elements in the column 624 of the SEM 620, which are not shown in FIG. 6. Furthermore, the computer system 670 comprises a setting unit 674 for setting and controlling various parameters of the SEM 620. The parameters that can be set by the setting unit 674 can be, for example, the magnification, the focus of the electron beam 628, one or more settings of the astigmatism corrector, the beam displacement, the position of the electron source, and / or one or more apertures (not shown in FIG. 6).

[0120] The scanning unit 672 and / or the setting unit 674 may for example perform or control or contribute to the inspection and / or processing of the mask 510 in the working area 560 by use of an embodiment of the method according to the invention.

[0121] Furthermore, the computer system 670 comprises a memory unit 676 in which instructions for carrying out, for example, one embodiment of the method according to the invention can be stored. The computer system 670 can include one or more processors designed to carry out such instructions, i.e. to control and activate corresponding components of the apparatus 600 (e.g. the SEM 620, the scanning unit 672, the setting unit 674 and / or a gas supply system not yet described) according to commands. The processor can include, for example, a powerful graphics processor.

[0122] 6 may be integrated into device 600 or may take the form of a dedicated device. Computer system 670 may be embodied using hardware, software, firmware, or a combination thereof.

[0123] To treat the defects 550 of the mask 510 and / or to write the (first and / or second) reference markings 540 and / or 580 on the mask 510, the apparatus 600 of FIG. 6 preferably comprises a number of different storage containers for different processes or precursor gases. In the apparatus 600 given as an example, two storage containers are shown. However, the apparatus 600 can also have three or more storage containers for treating the mask 510 and / or for writing the reference markings 540, 580 on the mask 510. The first storage container 652 stores a precursor gas or deposition gas, which can be used in cooperation with the electron beam 628 of the SEM 620, for example, to deposit a material for generating the reference markings 540, 580 of the mask 510. Furthermore, the electron beam 628 of the SEM 620 can be used, for example, to deposit a missing absorber material of one of the pattern elements of the mask 510. The second reservoir 662 contains an etching gas that can be used, for example, to etch the defect 550 .

[0124] Each storage vessel 652, 662 has its own valve 654 and 664, respectively, for controlling the amount of gas particles supplied per unit time or the gas flow rate at the point of incidence 635 of the electron beam 628 on the surface 520 of the mask 510. Furthermore, the two storage vessels 652, 662 have their own gas supply 656, 666, which terminates in a nozzle 658, 668 near the point of incidence 635 of the electron beam 628 on the mask 510. In the device 600 shown as an example in FIG. 6, the valves 654, 664 are integrated in the vicinity of the storage vessels. In an alternative embodiment, the valves 654, 664 can be located in the vicinity of the corresponding nozzles 658 and 668, respectively (not shown in FIG. 6). Each storage vessel 652, 662 can have its own elements for individual temperature setting and control. The temperature setting allows both cooling and heating of each precursor gas. In addition, gas supplies 656, 666 may each similarly have their own elements for setting and monitoring the temperature at which each precursor gas is delivered at the reaction site (also not shown in FIG. 6).

[0125] The apparatus 600 of FIG. 6 may comprise a pumping system for generating and maintaining the required vacuum. The pumping system is not shown in FIG. 6 for clarity. In addition, the apparatus 600 may comprise a suction extraction device (also not shown in FIG. 6). The pump or the suction extraction device in combination with the pumping system allows fragments or components generated during decomposition of the precursor gases, which are not required for the local chemical reaction, to be extracted from the vacuum chamber 602 of the apparatus 600 substantially at the point of generation. Unwanted gas components are pumped out of the vacuum chamber 602 locally at the position of incidence of the electron beam 628 on the mask 510 before they can disperse and settle in the vacuum chamber 602 of the apparatus 600, thus preventing contamination of the vacuum chamber 602. In the following, further embodiments of the invention are described.

[0126] Embodiment 1 relates to a method for generating correction marks on a lithographic object using a particle beam, in particular for calibrating the operation, the method comprising: a. generating a first group of calibration marks (A1, B1, C1); b. generating a second group of correction marks (A2, B2, C2); Including, c. A separation of the correction marks in the first group and in the second group is less than a separation between the correction marks from the first group and the correction marks from the second group.

[0127] Embodiment 2: A method according to embodiment 1, wherein the first group and / or the second group include correction marks having at least partially the same form.

[0128] Embodiment 3: A method according to any of embodiments 1-2, wherein at least one correction mark from the first group and / or the second group is composed of a plurality of geometric shapes.

[0129] Embodiment 4: The method according to any one of embodiments 1 to 3, wherein the number of correction marks from the first group and / or the second group is at least 3, preferably at least 4.

[0130] Embodiment 5: A method according to any one of embodiments 1 to 4, wherein the separation of the correction marks within the first group and / or within the second group is smaller than the separation between the correction marks from the first group and the correction marks from the second group, being at least 5 times smaller, preferably at least 10 times smaller, particularly preferably at least 20 times smaller.

[0131] Embodiment 6: The method according to any one of embodiments 1 to 5, wherein the production is based at least in part on a particle beam induced deposition process and / or a particle beam induced etching process.

[0132] Embodiment 7: The method according to any one of embodiments 1 to 6, further comprising generating at least one third group of correction marks (A3, B3, C3), wherein the separation of the correction marks within an individual group is smaller than the separation between correction marks from two different groups.

[0133] Embodiment 8: A method according to embodiment 7, in which the generated groups surround the working area such that a connecting line between two correction marks of different groups in each case can surround the working area of ​​the operation (500) without crossing the working area.

[0134] A ninth embodiment relates to a method for calibrating a motion on a lithographic object using at least one local group of spaced apart correction marks (A1, B1, C1; A2, B2, C2) using a particle beam, the method comprising the following sequence: S1. Selecting a sequence tuple (A1;A2), where the sequence tuple includes a subset of correction marks of at least one group; S2. performing a calibration based at least in part on the sequence tuple (A1;A2); S3. Performing at least some of the actions based at least in part on the performed calibration; and The present invention relates to a method comprising the steps of:

[0135] Embodiment 10: The method according to embodiment 9, wherein the selection of the sequence tuple is based at least in part on an evaluation of a predetermined criterion associated with at least one correction mark of at least one group.

[0136] Embodiment 11: A method according to embodiment 10, wherein the predetermined criteria comprises at least one of the following criteria: a degree of wear of the at least one correction mark, a contrast of the at least one correction mark, a gradient image of the at least one correction mark, an autocorrelation function of an image of the at least one correction mark, a cross-correlation function of at least two images of the at least one correction mark.

[0137] Embodiment 12: A method according to any one of embodiments 9 to 11, wherein the sequence is repeated at least once and in the process at least two sequence tuples comprising different subsets of correction marks are selected.

[0138] Embodiment 13: A method according to any one of embodiments 9 to 12, wherein the sequence tuple is selected to include a subset of correction marks from each of at least m groups of correction marks, where m is 2 or more, and the separation of correction marks within an individual group is less than the separation between correction marks from two different groups.

[0139]

[0036] Embodiment 14: The method according to embodiment 13, wherein the selection of the sequence tuple is based at least in part on the expected number of sequences in operation.

[0140] Embodiment 15: The method according to any one of embodiments 9 to 14, wherein the selection of sequence tuples over the sequence of operations is performed according to a predefined order.

[0141] Embodiment 16: The method according to any one of embodiments 9 to 15, wherein the selection is performed according to a cyclic order of the sequence tuples, a randomized order of the sequence tuples, and / or according to rows of the same sequence tuple.

[0142] Embodiment 17: A method comprising: determining a transformation of particle beam parameters associated with the two sequence tuples; performing a calibration based at least in part on the determined transformation; and The method according to any one of embodiments 9 to 16, further comprising:

[0143] Embodiment 18: Determining the conversion comprises: determining relative positions of correction marks of different sequence tuples with respect to each other; Determining relative positions of correction marks of different sequence tuples with respect to one or more structures of the lithography object; 18. The method according to embodiment 17, comprising at least one of:

[0144] Embodiment 19: A method according to any one of embodiments 1 to 18, wherein the calibration includes determining drift of the particle beam and / or correcting drift of the particle beam.

[0145] Embodiment 20: The method according to any one of embodiments 1 to 19, wherein the operation includes repairing the defect.

[0146] Embodiment 21 is an apparatus for generating correction marks on a lithographic object and / or calibrating operations using a particle beam, comprising: a. Means for carrying out the method according to any one of claims 1 to 20; b. Means for executing a computer program; The present invention relates to an apparatus comprising:

[0147] Embodiment 22 relates to a computer program comprising instructions which, when executed, cause an apparatus according to embodiment 21 to perform the method steps of the method according to any one of embodiments 1 to 20.

[0148] Embodiment 23: An apparatus according to embodiment 21 having a memory containing a computer program according to embodiment 22.

Claims

1. A method for calibrating a motion on a lithographic object using at least one local group of spaced apart correction marks (A1, B1, C1; A2, B2, C2) using a particle beam, comprising the steps of: S1. Selecting a sequence tuple (A1; A2), the sequence tuple including a subset of the correction marks of the at least one group; S2. performing a calibration based at least in part on said sequence tuple (A1;A2); S3. performing at least some of said operations based at least in part on said performed calibration; Including, The method of claim 1, wherein the sequence is repeated at least once and at least two sequence tuples are selected that include different subsets of correction marks.

2. The method of claim 1 , wherein the selection of the sequence tuples is based at least in part on an evaluation of a predetermined criterion associated with at least one correction mark of the at least one group.

3. The predetermined criteria are the following criteria:

3. The method of claim 2, comprising at least one of a wear degree of the at least one correction mark, a contrast of the at least one correction mark, a gradient image of the at least one correction mark, an autocorrelation function of an image of the at least one correction mark, a cross-correlation function of at least two images of the at least one correction mark.

4. 4. The method of claim 1, wherein the sequence tuples are selected to include a subset of correction marks from each of at least m groups of correction marks, where m is 2 or more, and a separation of the correction marks within an individual group is less than a separation across correction marks from two different groups.

5. The method of claim 4 , wherein the selection of the sequence tuples is based at least in part on an expected number of sequences in the run.

6. The method of any one of claims 1 to 5, wherein the selection of the sequence tuples over the sequence of operations is performed according to a predefined order.

7. The method according to any one of claims 1 to 6, wherein the selection is performed according to a cyclic order of sequence tuples, a randomized order of sequence tuples, and / or according to rows of the same sequence tuple.

8. determining a transformation of particle beam parameters associated with the two sequence tuples; performing the calibration based at least in part on the determined transformation; and The method of any one of claims 1 to 7, further comprising:

9. Determining the transformation comprises: determining relative positions of correction marks of different sequence tuples with respect to each other; determining the relative positions of the correction marks of the different sequence tuples with respect to one or more structures of the lithography object; The method of claim 8 , comprising at least one of:

10. A method for generating correction marks on a lithographic object using a particle beam, in particular for calibrating the operation, comprising: a. generating a first group of correction marks (A1, B1, C1); b. generating a second group of correction marks (A2, B2, C2); Including, c. a separation of the correction marks within the first group and within the second group is less than a separation between a correction mark from the first group and a correction mark from the second group; a separation of the correction marks within the first group and / or within the second group is at least 5 times smaller, preferably at least 10 times smaller, particularly preferably at least 20 times smaller, than a separation between a correction mark from the first group and a correction mark from the second group.

11. The method according to claim 10 , wherein the first group and / or the second group comprise correction marks having at least partly the same form.

12. The method according to claim 10 or 11, wherein at least one correction mark from the first group and / or the second group is comprised of a plurality of geometric shapes.

13. The method according to any one of claims 10 to 12, wherein the number of correction marks from the first group and / or the second group is at least three, preferably at least four.

14. The method according to any one of claims 10 to 13, wherein said generating is at least partly based on a particle beam induced deposition process and / or a particle beam induced etching process.

15. The method according to any one of claims 10 to 14, further comprising generating at least one third group (A3, B3, C3) of correction marks, wherein a separation of the correction marks within an individual group is smaller than a separation between correction marks from two different groups.

16. 16. The method of claim 15, wherein the generated groups surround a working area (500) of an operation such that a connecting line between two correction marks of different groups can surround the working area without crossing the working area.

17. The method according to any one of the preceding claims, wherein the calibration comprises determining a drift of the particle beam and / or correcting a drift of the particle beam.

18. The method of any one of claims 1 to 17, wherein the action comprises repairing a defect.

19. 1. An apparatus for generating correction marks on a lithographic object and / or calibrating operations using a particle beam, comprising: a. means for carrying out the method of any one of claims 1 to 18; b. Means for executing a computer program; An apparatus comprising:

20. A computer program comprising instructions which, when executed, cause an apparatus according to claim 19 to perform the method steps of the method according to any one of claims 1 to 18.

21. 20. Apparatus according to claim 19, comprising a memory containing a computer program according to claim 20.

Citation Information

Patent Citations

  • Dynamic creation of backup fiducials

    JP2017021006A