DE102024104438A1 - Method, device and computer program for determining and correcting a runout error when rotating a sample - Google Patents

Abstract

The present application relates to a method (800) for generating a calibration structure (300) for determining a runout error when rotating a sample (700) about a rotation axis (180), comprising the steps of: (a) generating (820) a first marking (310) on a surface (380) of a substrate (370), on which the calibration structure (300) is to be produced, at a predetermined distance (340) from the rotation axis (180); (b) rotating (830) the substrate (370) by a predetermined angle of rotation about the rotation axis (180); and (c) generating (840) a second marking (320-1 to 320-10) on the surface (380) of the substrate (370) for generating the calibration structure (300). Further aspects relate to determining the runout error and correcting it when rotating a sample (700).

Description

1. Technical area

The present invention relates to methods, devices, and computer programs for determining and correcting a runout error during the rotation of a sample. In particular, the present invention relates to methods for generating a calibration structure, methods for determining a runout error using a calibration structure, a computer program for determining and correcting a runout error, and a method and device for correcting a runout error during the rotation of a sample.

2. State of the art

As a result of the growing integration density in the semiconductor industry, photolithography masks must image increasingly smaller structures on wafers. To create the small structure dimensions imaged on the wafer, photolithographic masks or templates for nanoimprint lithography with ever smaller structures or pattern elements are required. The manufacturing process for photolithographic masks and templates for nanoimprint lithography is therefore becoming increasingly complex, time-consuming, and ultimately expensive. Due to the tiny structure sizes of the pattern elements of photolithographic masks or templates, errors in mask or template manufacturing cannot be ruled out. These must be repaired whenever possible.

Defects or errors in photolithographic masks, photomasks, exposure masks, or simply masks are often repaired by applying one or more process or precursor gases to the repair site and scanning or probing the defect, for example, with an electron beam. Typically, the electron beam induces a local chemical reaction, which, depending on the precursor gas used, leads to a local etching process, which can be used to remove excess material from the photomask or a template for nanoimprint lithography. Alternatively, in the presence of an appropriate precursor gas, the electron beam induces a local chemical deposition reaction, which locally deposits material on the photomask, thus replacing locally missing material from the mask.

Another cause of defects in photolithographic masks are particles that are created, for example, during mask handling and settle on the mask. These particles, which interfere with the image quality of the mask, must also be removed from the mask. Interfering particles can be removed from the photomask using a local particle beam-induced etching process. Furthermore, a micromanipulator, for example, in the form of a scanning probe microscope, can be used to remove excess material, such as particles present on the mask, from the photomask through interaction with the micromanipulator.

Due to the increasingly smaller structures of photomasks and the decreasing actinic wavelength at which masks are exposed, ever smaller defects and/or smaller particles are having a disruptive effect on the imaging behavior of photomasks. For example, in masks for the extreme ultraviolet (EUV) wavelength range, the actinic wavelength is in the range of approximately 10 nm to 15 nm. This means that increasingly better tools are needed to process defects in photolithographic masks. Furthermore, this development also results in increasing demands on the precision with which identified defects must be approached for repair.

Due to increasing precision requirements on the one hand and expanded movement capabilities of a sample stage on the other, the precise alignment or calibration of a mask to be repaired with respect to the sample stage or mask plate is becoming increasingly complex and time-consuming. Aligning or calibrating a mask to be repaired on a sample stage relative to the repair tool is therefore increasingly affecting the repair times for defective masks.

In particular, the ability to rotate a mask, or more generally a sample, around an axis of a sample stage leads to additional effort when aligning the mask or sample with a repair tool and to new sources of error inherent in the rotation process. This lengthens repair processes and reduces the throughput of masks to be repaired.

The present invention is therefore based on the problem of at least partially improving a turning process of a sample.

3. Summary of the invention

According to one embodiment of the present invention, this problem is at least partially solved by the subject matter of the independent claims of the present application. Exemplary embodiments are described in the dependent claims.

A first embodiment relates to a method for generating a calibration structure for determining a runout error when rotating a sample about a rotation axis, comprising the steps of: (a) generating at least one first mark on a surface of a substrate on which a calibration structure is to be generated, at a predetermined distance from the rotation axis; (b) rotating the substrate by a predetermined angle of rotation about the rotation axis; and (c) generating a second mark on the surface of the substrate to generate the calibration structure.

The inventors discovered that rotating a sample with a defect can lead to a change in the position of a repair tool at the defect's machining location. The change in the sample's position can be caused by a rotational axis whose position in a plane perpendicular to the rotational axis is not constant during or during rotation. This undesirable lateral or radial movement of the rotational axis is called runout, axial offset, or radial runout. When rotating a sample, the axial offset results in the machining location not matching the determined position of a defect to be machined. This can result in the tool machining a defect being incorrectly positioned laterally, i.e., in a plane perpendicular to the rotational axis. In the worst case, rotating the sample can inadvertently guide a repair tool to an incorrect, i.e., defect-free, location on the sample, and the defect machining can generate a new defect in the sample instead of correcting the existing one. An axis offset, i.e. a radial movement of the rotation axis, leads to a position error on the surface of a sample, which is referred to below as runout error.

A calibration structure described herein advantageously enables the determination of a concentricity error when rotating a sample without requiring the use of additional measuring devices. In particular, the methods described in this application enable the determination and correction of position errors when approaching a point on a sample, which involves rotating the sample, without having to break the vacuum of a repair tool. Rather, based on the determined concentricity error, the position change of a repair tool or a defect caused by rotating the sample can be determined and corrected. Furthermore, a defect processing point can be located at any position on the sample. Thus, a calibration structure described herein, in combination with the methods explained below, fully utilizes the defect repair possibilities improved by rotating the sample. Furthermore, these methods reliably prevent inadvertent processing of a sample at the wrong location.

Creating the calibration structure may comprise: repeating steps (b) and (c) to create a predetermined number of marks of the calibration structure.

The predetermined number of markings of the calibration structure can comprise at least 4, at least 8, at least 16, at least 32, or at least 64 markings. It is also possible to create a larger or much larger number of markings on a calibration structure. In particular, it is advantageous to adapt the number of predetermined markings to the required precision for determining the concentricity error.

The generation of the markings of the calibration structure can also be error-prone.

On the one hand, rotation by a specified angle of rotation can be subject to a rotation error, i.e., a specified, nominal angle of rotation differs from a measured angle of rotation. This difference can occur during the creation of the calibration structure and can advantageously be corrected when determining a radial runout error based on the generated calibration structure.

Secondly, even with ideal turning, the positions at which the at least two marks are created can vary slightly. In addition, the shape and form of the generated marks can be subject to small changes. By increasing the number of generated marks, the influence of these error sources on the accuracy of determining a runout error based on the manufactured calibration structure can be limited. However, generating a large number of marks for a calibration structure increases the experimental effort required for its production. By automating the mark generation process, however, the operational time and cost of a calibration structure can be limited.

The at least two markings can be created using various techniques. For example, they can be created by creating local depressions in the surface of the calibration structure, for example, with the tip of a micromanipulator. Alternatively, markings can be created by applying a substance with an ink to the surface of the calibration structure using a probe of a scanning probe microscope (SPM), such as an AFM (atomic force microscope) (dip pen nanolitho). graphy). It is also possible to generate the at least two markings by performing a particle beam-induced (e.g., with charged particles, e.g., electrons) deposition process and/or etching process on the surface of the calibration structure. To perform a particle beam-induced deposition process, at least one precursor gas in the form of a deposition gas is provided on the calibration structure. To perform a particle beam-induced etching process, at least one precursor gas in the form of an etching gas is provided on the calibration structure.

The (maximum) diameter of the at least two markings can each be <50 nm, preferably <30 nm, more preferably <15 nm, and most preferably <10 nm. The height or depth of the at least two markings of the calibration structure can be >10 nm, preferably >20 nm, more preferably >35 nm, and most preferably >50 nm, but in some examples less than 200 nm, less than 100 nm, or less than 75 nm.

Furthermore, it is advantageous if the at least two markings stand out clearly from the surface of the calibration structure during imaging. Simple geometric shapes, such as a cylinder, are therefore preferred, so that their positions can be determined with great precision. Furthermore, it is advantageous if the markings, in addition to a topographic contrast, also exhibit a material contrast with the material of the substrate of the calibration structure if they are imaged by scanning with a charged particle beam.

The material of the at least two markers may comprise carbon, oxygen and a metal, such as molybdenum.

The specified angle of rotation for generating the markings can be kept constant each time step b) is repeated. By generating the at least two markings after rotating the substrate of the calibration structure by the same angle of rotation, the effort required for evaluating and determining the runout error can be minimized. Furthermore, the angular error typically depends on the size of the angle of rotation. In mechanical systems, small angles of rotation usually lead to small angular errors.

However, it is also possible to change the angle of rotation so that the specified angle of rotation can change between the individual repetitions of step (b).

A sum of the predetermined angles of rotation for generating the at least two markings can be at least 22.5°, preferably at least 45°, at least 180°, at least 270° or even at least 359°.

It is advantageous to create calibration structures whose markings extend over the largest possible angular range around their rotational axis. This allows for the determination of a radial runout error with high precision across the entire range of the rotational axis's angle.

The predetermined distance between the rotation axis and the first marking can be the distance at which the (N-1) further markings are created. However, it is also possible for the predetermined distance to be changed by moving the substrate before the second and/or at least one of the further (N-1) markings is created. Changing the predetermined distance from the rotation axis depending on the predetermined angle of rotation (or leaving it constant) can create the predetermined number of markings essentially in the shape of a predetermined geometric figure. For example, the predetermined distance from the rotation axis can increase or decrease linearly with the angle of rotation, thereby creating the predetermined number of markings on a spiral line (if left constant, a circular shape would ideally be expected). This allows two or more markings to be created at different distances from the rotation axis for one angle of rotation, so that the sum of the angles of rotation can, for example, be a multiple of 360°. This makes it possible, on the one hand, to reduce an angular error when determining the position of the generated markings and, on the other hand, to reduce the position error of the various markings assigned to a given angle of rotation.

The term “substantially” means here, as elsewhere in this application, “within currently typical design, measurement and manufacturing tolerances”.

If the specified distance from the rotation axis is independent of the rotation angle (and the substrate is merely rotated), the specified number of marks is essentially generated on a circular arc around the rotation axis. In particular, generating a specified number of marks on two circular arcs with different radii allows for a precise analysis of the error sources discussed above.

The rotation axis can be oriented essentially perpendicular to the surface of the calibration structure. Under this boundary condition, an axis offset or Radial run-out and axial run-out can occur. In the case of a radial run-out, the rotational axis moves in a plane oriented perpendicular to the rotational axis, i.e., radially to the surface of the calibration structure. In the case of axial run-out, the surface of the calibration structure moves as a whole along the rotational axis during rotation. If the rotational axis is not oriented parallel to the normal of the surface of the calibration structure, the surface normal performs a wobbling motion around the rotational or rotary axis during rotation. The latter wobbling problem is addressed by the patent specification DE 10 2020 209 638 B 3 and the patent application DE 10 2023 205 623.2 dedicated to the applicant.

The surface of the substrate on which the at least two marks are produced may be substantially planar.

A flat surface of a calibration structure allows a precise determination of the positions of the generated at least two markings, since their focal plane does not change essentially during imaging.

Creating the calibration structure may include aligning the rotational axis of the substrate of the calibration structure with respect to an axis of a tool and creating a central mark at a position of the rotational axis on the surface of the substrate.

Aligning the rotation axis of the substrate of the calibration structure with respect to the axis of the tool for generating the at least two markings or the predetermined number of markings may comprise: determining coordinates of the rotation axis on the surface of the substrate of the calibration structure in a coordinate system connected to the calibration structure.

Determining the coordinates of the rotation axis may comprise: (a) measuring a first set of coordinates for at least two marking elements present on the substrate of the calibration structure without rotating the substrate of the calibration structure; (b) rotating the substrate of the calibration structure by an angle 0° < α < 180° or 180° < α < 360°; and measuring a second set of coordinates for the at least two marking elements present on the substrate of the calibration structure. Measuring the marking elements present on the substrate of the calibration structure may be performed, for example, by scanning over the marking elements with the focused particle beam. Alternatively or additionally, measuring the marking elements may be performed by scanning with a probe or a measuring tip of a scanning probe microscope (SPM) and/or by optical imaging.

Determining the coordinates of the rotation axis may comprise: determining the coordinates of the rotation axis from the first and second sets of measured coordinates of the at least two marking elements present on the calibration structure.

The marking elements present on the calibration structure can be applied to its surface during the calibration structure's manufacture. For example, a calibration structure in the form of a photomask or a mask blank typically has a number of marking elements (fiducial marks) that are typically applied at regular intervals on the side of the mask structured by the pattern or the side of a mask blank to be structured. Each of the marking elements usually has a reference point. The mask coordinates (u, v) specified by mask manufacturers or mask blank manufacturers refer to this point. The coordinates (u _D , v _D ) of a defect in a sample in the form of a mask or mask blank are also related to these reference points. To maximize the measurement accuracy when measuring the two reference points of the marking elements, it is advantageous to use marking elements that are as far apart as possible from each other on the calibration structure, which is in the form of a mask or mask blank.

Based on the two measured sets of coordinates of two marking elements, where the calibration structure is rotated before measuring the second set, the coordinates of the rotation axis of the calibration structure can be determined in a coordinate system connected to the calibration structure.

Determining the coordinates of the rotation axis of the calibration structure may comprise: determining the coordinates of the rotation axis from the first and second sets of measured coordinates of the at least two marking elements.

By selecting a rotation angle in the range of 90° to determine the coordinates of the rotation axis of the calibration structure, the accuracy of determining the rotation axis or the coordinates of the rotation axis can be optimized. It is therefore advantageous to select a rotation angle in the range of 90° for rotating the calibration structure to determine the position of the rotation axis on the calibration structure.

The measurement of the at least two marking elements can be carried out, for example, by scanning them with a focused particle beam and/or a probe of an SPM.

The electron beam of a scanning electron microscope (SEM) or a modified SEM is often used to measure the fiducial marks mentioned above, or their reference points. An electron beam can be focused on a very small spot (D _S < 1 nm). This enables very high lateral resolution when determining the reference points of the marking elements of the calibration structure. Furthermore, imaging the marking elements with an electron beam causes little or no damage to the marking elements and thus to the calibration structure as a whole.

After determining the coordinates of the rotation axis on a sample, the coordinates of a defect (u _D , v _D ), typically specified in sample coordinates, can be converted into the coordinate system of the sample stage (x _D , y _D , z _D ). The sample stage can move the sample placed on it to a designated repair location of the defect. Furthermore, a repair tool can be positioned with high precision to machine the defect.

A particle beam may comprise at least one element from the group: a photon beam, an electron beam, an ion beam, an atom beam, or a molecular beam.

Creating the calibration structure may include: creating a central mark at a position of the rotation axis on the surface of the substrate of the calibration structure.

By aligning the coordinates of the rotational axis of the substrate of the calibration structure with the axis of the tool used to create the at least two markings on the substrate of the calibration structure, a central marking establishes a reference point on the substrate of the calibration structure, to which the positions of the at least two markings can be related during production of the calibration structure. To mark this point on the substrate of the calibration structure, a central marking can be created on the substrate of the calibration structure, for example, after aligning the beam axis of the particle beam or the probe of an AFM, which is used to create the at least two markings, with the rotational axis of the calibration structure.

The calibration structure and/or a sample can be arranged on a sample stage. The sample stage can be displaceable along at least one axis lying in a plane substantially perpendicular to the rotational axis of the calibration structure. Furthermore, the rotational axis of the sample stage can be substantially parallel to the axis of the tool used to create the markings on the surface of the calibration structure. Adhering to this condition increases the positioning accuracy when creating the at least two markings for generating a calibration structure. If the markings are created, for example, using a scanning particle microscope and providing a precursor gas, its axis corresponds to the beam direction of the particle beam. If a scanning probe microscope (SPM) is used to create the markings, which, for example, creates depressions in the substrate of a calibration structure or deposits ink on the substrate of a calibration structure, its axis is parallel to the direction of the probe or measuring tip of the SPM.

The sample stage can be the sample stage of a repair tool. In particular, the repair tool can induce a local particle beam-induced etching and/or deposition process. Alternatively and/or additionally, the repair tool can comprise one or more SPMs, which can be used as manipulators or micromanipulators for processing a sample, for example, a defect in the sample. Furthermore, the one or more SPMs can be used to generate and/or image the at least two markings of a calibration structure.

The first and second markers can be created using the tool that creates the central marker.

Creating the calibration structure may include: shifting the substrate of the calibration structure by a vector $\overrightarrow{r},$ which is parallel to a plane oriented perpendicular to the axis of rotation of the substrate of the calibration structure, and creating a first mark on the surface of the substrate of the calibration structure.

By shifting the substrate of the calibration structure by a vector $\overrightarrow{r}$ A predetermined distance is set between the rotation axis and a first mark on the surface of the substrate of the calibration structure. Furthermore, the end point of the vector $\overrightarrow{r}$ the position or the new coordinates of the shifted rotation axis on the substrate of the calibration structure. Furthermore, the direction of the shift can be used to determine the reference point for the rotation angle of the rotation axis. A runout error dependent on the angle of rotation can be obtained from this reference point.

The magnitude of the vector $\overrightarrow{r}$ can range from 0.2 µm to 2000 µm, preferably from 0.5 µm to 500 µm, more preferably from 1 µm to 100 µm, and most preferably from 3 µm to 20 µm. These numerical values represent a compromise between the effort required to image the generated calibration structure on the one hand and the achievable accuracy in determining a runout error on the other. It is of course possible to make the best possible use of the entire surface of the substrate on which the calibration structure is generated to generate the calibration structure.

The larger the magnitude of the vector $\overrightarrow{r}$ The larger the specified distance between the rotation axis and the first marking, the more precisely a specified angle of rotation and the positions of the at least two markings can be determined from an image of the calibration structure. From this perspective, the largest possible specified distance between the rotation axis and the at least two markings is advantageous. However, this statement only applies without restriction to optical imaging of the markings of the calibration structure in a single image. If, on the other hand, the calibration structure is scanned with the probe of an AFM or the electron beam of an SEM, an image of the markings of the calibration structure must be composed from different scanning processes if their distance exceeds the maximum scan range of the respective imaging tool. However, composing an image from the data of several different scanning processes is error-prone. Depending on the imaging tool, a compromise must therefore be found between the specified distance of the markings from the rotation axis on the one hand and the size of the scan range of the imaging tool on the other. When using an SEM to image the calibration structure, specified distances in the range of 10 µm have proven advantageous. Under these conditions, the markings of the calibration structure can be scanned in a raster process.

The generation of the at least two markings may comprise: rotating the displaced substrate of the calibration structure by a predetermined angle of rotation about the axis of rotation and generating an i-th marking of the predetermined number of markings.

The predetermined angle of rotation may range from 1° to 150°, preferably 1° to 50°, preferably 1° to 20°, and most preferably 2° to 10°.

The method for generating the calibration structure may further comprise: shifting the calibration structure by a vector $-\overrightarrow{r}$ after creating a last of the at least two markers. After shifting the calibration structure by the vector $-\overrightarrow{r}$ The position of the rotation axis again essentially coincides with the position of the central marking. Furthermore, the rotation axis of the generated calibration structure again essentially coincides with the axis of the tool that created the at least two markings.

Furthermore, the method for generating the calibration structure may comprise: imaging the at least two generated markings of the calibration structure.

As already explained above, the imaging of the at least two markings of the calibration structure can be performed by scanning with the probe of an SPM, for example an AFM, and/or by scanning with a focused particle beam, such as an electron beam of an SEM. Furthermore, it is possible to optically image the at least two markings of the calibration structure. The latter can be performed by capturing an image with a camera or by scanning the markings using a photon beam. In this case, it is advantageous to use photons with short wavelengths, for example in the ultraviolet (UV) or deep ultraviolet (DUV) wavelength range, to increase the resolution.

A second embodiment relates to a method for determining a runout error when rotating a sample about a rotation axis, comprising the steps of: (a) obtaining at least one image of a calibration structure having at least two markings whose positions can be converted into one another by rotating the calibration structure by a predetermined angle of rotation about the rotation axis; and (b) determining the runout error based on the at least one image.

To determine a runout error resulting from an axis offset, a calibration structure can be manufactured - as described above - and based on its image, the runout error of a rotational axis of a sample stage can then be determined, which held the calibration structure during its manufacture - as explained above.

Obtaining the at least one image of the calibration structure may involve actually imaging the calibration structure. Alternatively or additionally, the image may be received, e.g., from a memory and/or via an interface.

Obtaining the at least one image of the calibration structure may, in some cases, comprise at least one of: obtaining the at least one image from a non-volatile memory, obtaining the at least one image via a network connection, or imaging the calibration structure.

It should be noted that the calibration structure according to the second embodiment may include a calibration structure according to the first embodiment.

A third embodiment relates to a method for determining a runout error when rotating a sample about a rotation axis, comprising the steps of: (a) imaging a calibration structure having at least one marking; (b) rotating the calibration structure by a predetermined angle of rotation; (c) imaging the rotated calibration structure; and (d) determining the runout error from the at least two images of the calibration structure.

If a sample has one or more structural elements, such as one or more pattern elements of a photomask, these can be used to determine the runout of a rotational axis of the sample stage used to position the sample. This eliminates the effort required to create a calibration structure.

The calibration structure according to the third embodiment may comprise a calibration structure according to the first and/or second embodiment.

Determining a runout error according to the second and/or third embodiment may include, for example: determining a look-up table for an x/y correction value for multiple rotation angles of a sample holder, determining a function of an x-y correction value as a function of a rotation angle of a sample holder, etc.

The calibration structure can comprise a sample having at least two markings in the form of marking elements with known position data. Such marking elements can be generated by the sample manufacturer, for example, in the case of a photomask, and their position data can be supplied or communicated together with the photomask.

Furthermore, the method for determining a runout error when rotating a sample about a rotation axis may comprise: aligning the rotation axis of the sample with one of the at least one marking. The at least one marking may comprise at least one marking element of the sample.

If a sample already has at least one marking, e.g., in the form of a marking element, such as fiducial marks, the sample itself can be used to determine the runout of a rotary axis of the sample stage, which positions the sample for processing relative to, for example, a repair tool. Since the position data of a reference point of a marking element, e.g., a photomask, is usually supplied with high precision by the sample manufacturer, precisely manufactured marking elements can facilitate the automation of the runout determination, for example, through the use of image recognition software.

The conversion of position data from a coordinate system connected to the sample into a coordinate system associated with the sample stage is outlined above. Details can be found in the patent specification DE 10 2020 209 638 B from the applicant.

The angle of rotation for rotating the calibration structure to generate images of the rotated calibration structure can correspond to the specified angle of rotation when rotating the substrate to generate the calibration structure. The number of images of rotated calibration structures can correspond to the number of specified markings of the calibration structure.

Obtaining the at least two images of the calibration structure having at least one marking may comprise at least one of: obtaining the at least two images of the calibration structure from a non-volatile memory, obtaining the at least two images of the calibration structure via a network connection, or imaging the non-rotated calibration structure and imaging the rotated calibration structure.

Determining the runout error of the rotation axis may include: determining position data of the at least two markings in the at least one image of the calibration structure, and/or determining position data of the at least one marking in the at least two images of the calibration structure. The calibration structure may have at least two markings with known position data.

The at least two images of the calibration structure having at least one marking may comprise: imaging the non-rotated calibration structure to generate a first image and imaging the rotated calibration structure to generate a second image.

The position data of the at least one image of the calibration structure, which has at least two markings, and the at least two images of the calibration structure, which has at least one marking, typically relate to a coordinate system associated with the calibration structure. Furthermore, the position data of the at least two marking elements of a sample relate to a coordinate system assigned to the sample. To determine the position of the rotation axis on the sample, the measured position data of the at least one marking element can be fitted to a circular arc. For this purpose, an iterative Levenberg-Marquardt algorithm can be used, for example. If the position of the rotation axis on the sample is known, the determined position data of the non-rotated marking element, the marking element rotated once, and the marking element rotated at least a second time can be related to nominal position data that describe rotation on a circular arc.

Determining the runout error of the rotation axis may comprise: determining deviations of the position data from nominal coordinates for each of the at least two markings of the calibration structure, and/or determining deviations of the position data from nominal coordinates of the at least one marking of the at least two images.

The nominal coordinates of the at least two markings of the calibration structure are the coordinates generated when point-like markings are created on the substrate of a calibration structure and when a rotation is performed in the mathematical sense on the substrate of the calibration structure. If the specified distance between the rotation axis and the first marking remains unchanged, the nominal coordinates of the first and the at least one second marking are generated on a circular arc.

When rotating one or more marking elements or structural elements of a sample, nominal coordinates are the coordinates resulting from performing at least one rotation in the mathematical sense of the marking element or structural element. The reference point of at least one marking element or structural element is considered to be point-like.

The error types and sources discussed above in the context of creating the calibration structure overlap in the determined deviations of the position data from nominal coordinates. In the following, it is assumed that the number of markings is large (N > 20) and that these largely cover the rotation angle range from 0° to 360° (θ _S > 340°). This at least partially compensates for the errors in creating the markings. Since the reference points of the marking elements present on a sample are known, the error contributions associated with creating a calibration structure are eliminated. It is therefore assumed below that the determined deviations are primarily caused by the rotation process of the calibration structure and/or the sample.

The detected deviations can include concentricity errors caused by an axis offset of the rotation axis and roundness errors caused by an out-of-roundness or eccentricity of the rotation axis of the sample stage.

The method in question does not separate the error contributions of the various error sources, but rather determines a combined error. If the determined deviations do not exhibit a roundness error, the concentricity errors, at a constant, specified distance, result in a circular arrangement of the markings of the calibration structure around the axis of rotation with a diameter that deviates from the diameter of the circle of the nominal arrangement of the markings. Due to the axial offset of the sample stage's axis of rotation, the experimentally determined radius is larger than the nominal circle radius when the axis of rotation rotates by 360°. The offset of the centers of the nominal and determined circles can be determined by fitting a circle to the determined circular arc of at least two markings. The difference between the diameter of the fitted circle and the diameter of the nominal circle corresponds to twice the concentricity error.

The described method enables the determination of concentricity errors and roundness errors independently of angular position errors, even if the at least two markings are generated with a small radius (< 3 µm) around the rotation axis.

The method may further comprise interpolating the runout error between the deviations of the i-th and the (i+1)-th marking to determine the runout error for rotation angles between those of the i-th and the (i+1)-th marking. The method may further comprise extrapolating between the N-th and the first marking of the calibration structure to determine the runout error for rotation angles between the N-th and the first marking.

Similar to the individual markings of a calibration structure, a individual rotations or rotation angles associated with the rotations of a sample rotation are interpolated or extrapolated to determine the runout error of the rotation axis of the sample table on which the sample is positioned.

By interpolating and/or extrapolating between two determined deviations, a table of the runout error as a function of the rotation angle of a sample stage's axis of rotation can be created, containing a scalable number of entries. In particular, interpolation and/or extrapolation allows the creation of a function that describes the runout error as a function of the rotation angle.

The interpolation and/or extrapolation may comprise a linear interpolation and/or a linear extrapolation between the i-th and (i+1)-th marking or the N-th and the first marking (or the i-th and the (i+1)-th rotation of the sample). The interpolation and/or extrapolation may comprise determining the curvature of a circular arc between the first and the N-th marking and interpolating and/or extrapolating between the i-th and the (i+1)-th marking (or the i-th and the (i+1)-th rotation of the sample) and the N-th and the first marking (or the N-th and the first rotation of the sample) along a circular arc with the determined curvature.

The method for determining a runout error may further comprise: determining a difference between a predetermined rotation angle and a nominal rotation angle and correcting the difference to determine the runout error of the rotation axis.

A specified angle of rotation may deviate from a nominal angle of rotation. This error causes the markings on a calibration structure to be generated at incorrect positions. This can degrade the determined angle dependence of the runout error. It is therefore necessary to determine the difference between a specified or measured angle of rotation and a nominal angle of rotation and to correct this error when determining the angle dependence of the runout error.

A predetermined angle of rotation of a sample can be determined by rotating a sample by this angle around a reference point of a first marking element and determining the position of the reference point of at least one second marking element before and after rotation. The difference can be determined by subtracting the measured angle of rotation from the nominal angle of rotation. In the table that shows the runout error as a function of the predetermined angle of rotation, the predetermined nominal angle of rotation can be corrected by the determined difference. The angle of rotation difference between the various markings of the calibration structure can be interpolated or extrapolated as described above. The same applies to the predetermined, nominal angles of rotation between the i-th and the (i+1)-th rotation of the sample for determining the runout error.

A fourth embodiment relates to a method for correcting a runout error during rotation of a sample, comprising the steps of: (a) rotating the sample by a predetermined angle of rotation about a rotation axis, wherein the angle of rotation is related to a predetermined orientation of the sample in a plane perpendicular to the rotation axis; and (b) correcting a runout error associated with the predetermined angle of rotation during rotation and/or at the end of rotation by translating the sample in a plane perpendicular to the rotation axis based on the angle of rotation.

By relating the sample's angle of rotation to a straight line in the plane perpendicular to the rotation axis, a runout correction table can be used, listing runout errors for a range of rotation angles. Alternatively, it is also possible to plot the runout of a sample stage's rotation axis as a function of the rotation angle. This allows the runout that occurs during sample rotation to be compensated for during rotation. As a result, a reference point on the sample is guided along a circular arc during rotation, and the rotation axis appears to exhibit no axial offset during rotation.

By correcting a runout error of a sample stage's rotational axis during or immediately after the rotation process, a method according to the invention enables a significant increase in the precision of a sample rotation process. As a result of the increased positioning accuracy when approaching a defect, the accuracy of a sample repair can be significantly improved.

The displacement of the sample in a plane perpendicular to the rotation axis can be carried out by a sample stage having at least two translation axes oriented in the plane perpendicular to the sample, which are not oriented parallel to each other.

A sample can be any sample that needs to be rotated for processing. For example, the sample can be a semiconductor component ment, such as an integrated circuit, a wafer, a photomask, or a template for nanoimprint lithography. A wafer may comprise a silicon wafer and/or a compound semiconductor wafer. A photomask may comprise a transmissive or a reflective photomask. A transmissive mask may include any conventional photomask, such as a binary mask, a phase-shifting mask, or a multiple-exposure mask. A reflective photomask may comprise a mask for the extreme ultraviolet (EUV) wavelength range, in particular a binary or a phase-shifting mask.

Finally, a fifth embodiment relates to a device for correcting a runout error during rotation of a sample, the device comprising: (a) means for rotating the sample by a predetermined angle of rotation about a rotation axis, wherein the angle of rotation is related to a predetermined orientation of the sample in a plane perpendicular to the rotation axis; and (b) means for correcting a runout error associated with the predetermined angle of rotation during rotation and/or at the end of rotation by means for translating the sample in a plane perpendicular to the rotation axis based on the angle of rotation.

The means for rotating the sample may comprise a sample stage with at least one axis of rotation. The means for displacing the sample may comprise a sample stage that is displaceable along at least two axes lying in a plane perpendicular to the axis of rotation, wherein the at least two axes are not parallel.

A device according to the invention may further comprise: means for imaging the sample while rotating the sample. The means for imaging the sample may comprise at least one element from the group: a camera, a CCD (charge coupled device) sensor, or a particle beam with an associated detector.

Since a device according to the invention allows not only the correction of a runout error of the rotational axis of the sample table but also an in-situ observation of a sample repair process, it enables a fast and reliable repair of a sample.

Furthermore, a device according to the invention can be configured to carry out the method steps according to one of the aspects described above.

A further embodiment relates to a device for generating a calibration structure for determining a runout error when rotating a sample about a rotation axis, comprising: (a) means for generating at least one first marking on a surface of a substrate on which a calibration structure is to be generated, at a predetermined distance from the rotation axis; (b) means for rotating the substrate by a predetermined angle of rotation about the rotation axis; and (c) means for generating a second marking on the surface of the substrate to generate the calibration structure.

Another embodiment relates to a device for determining a runout error during rotation of a sample, comprising: (a) means for obtaining at least one image of a calibration structure having at least two markings whose positions can be converted into one another by rotating the calibration structure by a predetermined angle of rotation about the axis of rotation; and (b) means for determining the runout error based on the at least one image.

Furthermore, another embodiment relates to a device for determining a runout error when rotating a sample about a rotation axis, comprising: (a) means for imaging a calibration structure having at least one marking; (b) means for rotating the calibration structure by a predetermined angle of rotation; (c) means for imaging the rotated calibration structure; and (d) means for determining the runout error from the at least two images of the calibration structure.

Furthermore, one aspect of the present invention relates to a computer program comprising instructions for carrying out the method steps described herein.

In principle, all aspects described herein with respect to method steps of the methods can be transferred to functionalities of a corresponding device and/or instructions of a corresponding computer program and vice versa.

For example, devices can be provided that have means for carrying out the methods described herein. These devices can, for example, be equipped with means for executing a computer program, wherein a computer program is stored in the memory of the device, which has instructions to automatically carry out the method steps described herein. For example, suitable calibration structures can be generated automatically, as described herein with reference to methods. In addition, suitable calibration structures can be imaged automatically, if necessary after appropriate rotation. Finally, calibration structures and/or images thereof can also be used automatically to determine a calibration function, e.g. to determine a rotation-angle-dependent correction value for an x- and/or y-position (e.g., a position along a direction perpendicular to the rotation axis) of a sample holder. Finally, devices can also be provided that automatically correct the x- and/or y-position of the sample holder upon rotation of the sample holder, so that essentially an ideal rotation can be achieved and an axis offset can be substantially suppressed.

4. Description of the drawings

• 1 eine Ansicht eines Probentisches darstellt, der drei Translationsachsen und eine Rotationsachse aufweist;
• 2 einen schematischen Schnitt durch einige Komponenten einer Vorrichtung wiedergibt, die das Erzeugen einer Kalibrierstruktur, das Abbildern der erzeugten Kalibrierstruktur, das Bestimmen eines Rundlauffehlers der Drehachse des Probentisches der Vorrichtung und das Korrigieren des Rundlauffehlers erlaubt;
• 3 eine Aufsicht auf eine beispielhafte Kalibrierstruktur zeigt;
• 4 im linken Teilbild eine Abbildung einer Kalibrierstruktur einer fehlerfreien Drehachse wiedergibt und im rechten Teilbild eine Abbildung einer Kalibrierstruktur zeigt, die durch Drehen um eine Drehachse eines Probentisches hergestellt wurde, die einen Rundheitsfehler aufweist;
• 5 im linken Teilbild, wie in 4, eine Abbildung einer Kalibrierstruktur einer idealen Drehachse darstellt und im rechten Teilbild eine Abbildung einer Kalibrierstruktur zeigt, die durch Drehen um eine Drehachse eines Probentisches hergestellt wurde, die einen Exzentrizitätsfehler aufweist;
• 6 ein Auflaufdiagramm zum Erzeugen einer Kalibrierstruktur illustriert, deren Abbildung eingesetzt wird, zum Bestimmen von Rundlauffehlern der Drehachse des Probentisches, der ein Substrat gehalten hat, auf dem die Kalibrierstruktur hergestellt wurde;
• 7 eine Probe in Form einer Maske oder eines Maskenrohlings darstellt, die drei Markierungselemente aufweist, deren Referenzpunkte zum Bestimmen von Rundlauffehlern einer Drehachse des die Probe haltenden Probentisches eingesetzt wird;
• 8 ein Flussdiagramm eines erfindungsgemäßen Verfahrens zum Erzeugen einer Kalibrierstruktur zum Bestimmen von Rundlauffehlern beim Drehen einer Probe wiedergibt;
• 9 ein Ablaufdiagramm eines ersten Verfahrens zum Bestimmen von Rundlauffehlern beim Drehen einer Probe um eine Drehachse eines Probentisches darstellt;
• 10 ein Ablaufdiagram eines zweiten Verfahrens zum Bestimmen von Rundlauffehlern beim Drehen einer Probe um eine Drehachse eines Probentisches wiedergibt; und
• 11 ein Flussdiagram zum Korrigieren von Rundlauffehlern beim Drehen einer Probe um eine Drehachse eines Probentisches präsentiert.

In the following detailed description, currently preferred embodiments of the invention are described with reference to the drawings, wherein

• 1 shows a view of a sample stage having three translation axes and one rotation axis;
• 2 shows a schematic section through some components of a device which allows the creation of a calibration structure, the imaging of the created calibration structure, the determination of a runout error of the rotational axis of the sample table of the device and the correction of the runout error;
• 3 shows a top view of an exemplary calibration structure;
• 4 in the left part of the image, an image of a calibration structure of a defect-free axis of rotation is shown, and in the right part of the image, an image of a calibration structure produced by rotating around an axis of rotation of a sample stage, which has a roundness error;
• 5 in the left part of the image, as in 4 , represents an image of a calibration structure of an ideal rotation axis and, in the right-hand part, shows an image of a calibration structure produced by rotating around a rotation axis of a sample stage, which has an eccentricity error;
• 6 illustrates a run-up diagram for creating a calibration structure, the image of which is used to determine run-out errors of the rotational axis of the sample stage that held a substrate on which the calibration structure was fabricated;
• 7 represents a sample in the form of a mask or a mask blank having three marking elements whose reference points are used to determine runout errors of a rotational axis of the sample stage holding the sample;
• 8 a flowchart of a method according to the invention for generating a calibration structure for determining runout errors when rotating a sample;
• 9 a flowchart of a first method for determining runout errors when rotating a sample about a rotation axis of a sample stage;
• 10 a flowchart of a second method for determining runout errors when rotating a sample about a rotation axis of a sample table; and
• 11 A flowchart for correcting runout errors when rotating a sample around a rotation axis of a sample stage is presented.

5. Detailed description of preferred embodiments

Currently preferred embodiments of the methods and apparatus according to the invention are explained below. The methods according to the invention are explained in detail using the example of generating and imaging a calibration structure for a sample stage with a rotational axis, as well as determining and correcting a runout error of the sample stage's rotational axis. The correction of a runout error is described, as an example, by rotating a photomask on a sample stage for its correction. However, the methods specified herein are not limited to rotating a photomask. Rather, they can be used to rotate any sample, such as a wafer, a semiconductor component, an integrated circuit, a template for nanoimprint lithography, and/or any microsystem technology component, during its manufacturing or repair processes. Furthermore, the use of the methods according to the invention is not limited to sample stages that have only one rotational axis. Naturally, they can also be used to determine and correct the runout errors of sample stages that have two or more rotational axes.

The 1 shows a view of an exemplary sample stage 100 for holding a photomask, which has three translational and one rotational or rotary axes. The translational axes are mutually perpendicular to each other and thus form an orthogonal coordinate system. This is illustrated below the sample stage 100. The sample stage 100 is also referred to below as mask stage 100 or stage 100. The base plate 110 of the sample stage 100 has rails 120 for displacing the sample receiving surface of the sample stage 100 in the y-direction. The carriage 130 of the sample stage 100 can be mounted on the Rails 120 are moved along the y-axis. On its upper side, the carriage 130 carries rails 140, which enable a second carriage 150 of the sample stage 100 to be moved along the x-direction. The second carriage 150 forms the base plate 160 for the displacement unit 170 in the z-direction. The rotation axis 180, which is oriented or aligned parallel to the z-direction, is arranged on the displacement unit 170 of the sample stage 100 for the z-direction 170. The rotation axis 180 carries the sample chuck 190, the sample holder 190, or the sample plate 190. In the following, the sample receiving surface 195 of the sample holder 190 refers to the sum of the points on which a sample rests on the sample holder 190.

The 2 shows a schematic section through some important components of an example of a device 200 designed or configured to produce a calibration structure. Furthermore, the device 200 can be used to image the markings of a calibration structure, for example, the previously produced calibration structure. Furthermore, the device 200 can rotate a sample in the form of a photomask several times, and a runout error of the rotation axis 180 of the sample stage 100 can be determined from the position data of the marking elements present on the mask. Finally, the device 200 can be configured to correct a rotation error of the rotation axis 180 during a rotation process and/or after its completion.

For this purpose, the device 200 has the sample table 100 of the 1 A sample 205, for example in the form of a photolithographic mask 205, can be arranged on this. The photomask 205 can have one or more defects in the form of excess material (“dark defects”) and/or missing material (“clear defects”). The defect(s) of the mask 205 are in the 2 not reproduced. The defect, or generally defects of excess or missing material, can be scanned and thus analyzed using a particle beam and/or a probe or a measuring probe of a scanning probe microscope (SPM) 280. Furthermore, defects can be corrected using a particle beam-induced machining process or machining with the probe of the SPM 280. For this purpose, the device 200 comprises a modified scanning particle microscope 210 in the form of a scanning electron microscope (SEM) 210.

In the SEM 210 of the 2 An electron gun 212 generates an electron beam 215 which is projected by the imaging elements arranged in the electron column 217, which are arranged in the 2 are not shown, is directed as a focused electron beam 215 onto the sample 205. The sample 205 is mounted on the sample stage 100 of the 1 A sample table 100 is also known in the field as a “stage.” As shown in the 2 symbolized by the arrows 207, the sample table 100 can, in addition to rotating about the rotation axis 180, move a sample 205 about three translation axes relative to the column 217 of the SEM 210. The movement of the sample table 100 can, for example, be carried out with the aid of micromanipulators which are arranged in the 2 are not shown. Thus, the sample stage 100 enables the positioning of the sample 205 for analyzing its defects by generating one or more images of the defect. For this purpose, the imaging elements of the column 217 of the SEM 210 can rasterize or scan the electron beam 215 over the sample 205. By rotating the four-axis sample stage 100, it enables the examination of one or more defects at different angles. The respective position of the translation axes of the sample stage 100 can be measured interferometrically (in the 2 (not shown). The sample stage 100 is controlled by signals from an adjustment unit 255. The adjustment unit 255 can be part of a computer system 230 of the device 200. Furthermore, moving the sample stage 100 in the beam direction allows the sample stage 100 to be lowered so that a sample or a calibration structure can be positioned under the electron beam 215 and the measuring probe of the SPM 280.

The device 200 may further comprise sensors that enable both a current state of the SEM 210 and the process environment in which the SEM 210 is used (such as a vacuum environment) to be characterized.

The electron beam 215 can further be used to induce a particle beam-induced machining process for correcting identified defects, for example, in the context of an electron beam-induced etching process (EBIE), for removing dark or opaque defects, generally excess defects, and/or for performing an electron beam-induced deposition process (EBID) for correcting clear defects, i.e., defects of missing material. Furthermore, the focused electron beam 215 can be used to generate markings on a calibration structure. In addition, in the device 200, the 2 the electron beam 215 can be used to analyze a repaired area of the sample 205 and to image the markings of a calibration structure.

The device 200 comprises a holding device 222 for holding a calibration structure 225. The calibration structure 225 can be described below in the context of 3 The holding device 222 may comprise the calibration structure 300 explained above. Furthermore, the holding device 222 may have a positioning unit 227. The positioning unit 227 enables the calibration structure 225, 300 to be positioned on the sample stage 100 below the electron beam 215 and/or below the probe or the measuring probe of the SPM 280.

The electrons backscattered from the electron beam 215 by the sample 205 or the calibration structure 225, 300 and the secondary electrons generated by the electron beam 215 in the sample 205 or the calibration structure 225, 300 are registered by the detector 220. The detector 220, which is arranged in the electron column 217, is referred to as an "in-lens detector." In various embodiments, the detector 220 can be installed in the column 217. The detector 220 is controlled by the setting unit 255 of the computer system 230 of the device 200.

The device 200 may include a second detector 221. The second detector 221 may be configured to detect electromagnetic radiation, particularly in the X-ray range. Thus, the second detector 221 enables the analysis of the material composition of markers deposited on the calibration structure 225, 300 as well as the sample 205. The detector 221 is also controlled by the adjustment unit 255.

The setting unit 255 of the computer system 230 can set the parameters of the electron beam 215 for inducing a deposition process and an etching process on the calibration structure 225, 300.

Furthermore, the computer system 230 of the device 200 can have a computing unit 240. The computing unit 240 receives the measurement data from the detector(s) 220, 221. The computing unit 240 can generate images in a grayscale representation or a gray value representation from the measurement data, for example from SE contrast data, which can be displayed on a monitor 232. Furthermore, the computer system 230 has an interface 237 via which the computer system 230 or the computing unit 240 can receive data from additional external detectors relating to the calibration structure 225, 300 or the sample 205. The received data can include images of the calibration structure 225, 300 and/or data from images of the sample 205 taken at different angles of rotation of the rotation axis 180 of the sample stage 100. Furthermore, the computer system 230 can transmit the measurement data from the detectors 220 and/or 221 to an external evaluation device via the interface 237. Furthermore, the computer system 230 of the device 200 can receive from an external evaluation device one or more processed or evaluated images or depictions of the calibration structure 225, 300 and/or one or more superimposed images of a sample 205, which were acquired at different angles of rotation of the rotation axis 180 of the sample stage 100.

The computer system 230 or the adjustment unit 250 can rotate the sample 205 by rotating the sample holder 190 about the rotation axis 180. In particular, the computer system 230 or the adjustment unit 250 can simultaneously translate the sample 205 held by the sample holder 190 of the sample stage 100 about two axes in the plane of the sample holder 190 and rotate it about the rotation axis 180 of the sample stage 100.

As already explained above, the electron beam 215 of the modified SEM 210 of the device 200 can be used to induce an electron beam-induced machining process, such as the calibration structure 225, 300 and/or the sample 205. To carry out these processes, the exemplary scanning electron microscope 210 of the device 200 of the 2 three different storage containers 250, 260 and 270.

The first reservoir 250 can store a first precursor gas in the form of a deposition gas, for example a metal carbonyl, such as chromium hexacarbonyl (Cr(CO) ₆ ), or a carbon-containing precursor gas, such as pyrene. With the aid of the precursor gas stored in the first reservoir 250, material can be deposited on the sample 205 or the calibration structure 225, 300 in a local chemical reaction, wherein the electron beam 215 of the SEM 210 acts as an energy supplier to split the precursor gas stored in the first reservoir 250, preferably into metal atoms and carbon monoxide molecules at the location where material is to be deposited, e.g., at the positions of the calibration structure 225, 300 where markings are to be created on the calibration structure 225, 300. This means that by the combined provision of an electron beam 215 and a precursor gas, an EBID process is carried out to create a calibration structure 225, 300.

In the in the 2 In the device 200 shown, the second reservoir 260 stores a precursor gas in the form of an etching gas, which enables the execution of a local electron beam induced etching process (EBIE). With the aid of an electron beam induced etching process, depressions can be created in the surface of a calibration structure. 225, 300. A precursor gas in the form of an etching gas can, for example, comprise xenon difluoride (XeF ₂ ), chlorine (Cl ₂ ), oxygen (O ₂ ), ozone (O ₃ ), water vapor (H ₂ O), hydrogen peroxide (H ₂ O ₂ ), nitrous oxide (N ₂ O), nitrogen monoxide (NO), nitrogen dioxide (NO ₂ ), nitric acid (HNO ₃ ), nitrosyl chloride (NOCl), ammonia (NH ₃ ) or sulfur hexafluoride (SF ₆ ) or a combination thereof.

The third reservoir 270 can store an additive gas that can be added as needed to the etching gas held in the second reservoir 260 or to the deposition gas stored in the first reservoir 250. Alternatively, the third reservoir 270 can store a precursor gas in the form of a second deposition gas or a second etching gas.

Each of the storage containers 250, 260 and 270 has in the 2 The scanning electron microscope 210 shown has its own control valves 252, 262, and 272 to control the amount of the corresponding gas provided per unit of time, i.e., the gas flow rate at the point where the electron beam 215 impinges on the sample 205 or the calibration structure 225, 300. The control valves 252, 262, and 272 are controlled by the adjustment unit 255. This allows the partial pressure ratios of the gas(es) provided at the processing location for carrying out an EBID and/or EBIE process to be adjusted within a wide range.

Furthermore, in the exemplary SEM 210, the 2 each reservoir 250, 260 and 270 has its own gas supply system 254, 264 and 274, which ends with a nozzle 256, 266 and 276 near the point of impact of the electron beam 215 on the sample 205 or the calibration structure 225, 300.

The storage containers 250, 260 and 270 can have their own temperature adjustment element and/or control element, which allows both cooling and heating of the respective storage containers 250, 260 and 270. This allows the storage and in particular the provision of the precursor gases of the deposition gas and/or the etching gas at the respective optimal temperature (in the 2 (not shown). The setting unit 255 can control the temperature setting elements and the temperature control elements of the storage vessels 250, 260, and 270. During the EBID and EBIE processing operations, the temperature setting elements of the storage vessels 250, 260, and 270 can also be used to adjust the vapor pressure of the process gas(es) stored therein by selecting an appropriate temperature.

The apparatus 200 may include more than one reservoir 250 for storing precursor gases of two or more deposition gases. Furthermore, the apparatus 200 may include more than one reservoir 260 for storing precursor gases of two or more etching gases.

The 2 The scanning electron microscope 210 shown can be operated under ambient conditions or in a vacuum chamber 242. To perform the EBID and EBIE processes, a negative pressure in the vacuum chamber 242 relative to the ambient pressure is necessary. For this purpose, the SEM 210 of the 2 a pump system 244 for generating and maintaining a required negative pressure in the vacuum chamber 242. When the control valves 252, 262 and 272 are closed, a residual gas pressure of < 10 ^-4 Pa is achieved in the vacuum chamber 242. The pump system 244 can comprise separate pump systems for the upper part of the vacuum chamber 242 for providing the electron beam 215 of the SEM 210 and the lower part 248 or the reaction chamber 248 (in the 2 not shown).

The device 200 of the 2 The SEM 210 presented has a single electron beam 215. However, it is also possible for the SEM 210 to have a source for a second particle beam. The second particle beam may comprise a photon beam and/or an ion beam (in the 2 not shown). Furthermore, the SEM 210 may have two or more electron beams 215 to perform two or more particle beam-induced machining processes or two or more measurement processes in parallel.

In addition, the 2 The exemplary device 200 shown includes a scanning probe microscope (SPM) 280, which is embodied in the device 200 in the form of an atomic force microscope 280 or an atomic force microscope (AFM) 280. The SPM 280 can be used to produce a calibration structure 225, 300 by generating markings in the form of depressions and/or to generate markings by depositing ink-containing substances at the locations on the calibration structure 225, 300 where markings are to be generated. Furthermore, the SPM 280 can be used to scan a calibration structure 225, 300 to generate an image of the calibration structure 225, 300. Furthermore, the SPM 280 can be used to repair the sample 205. For this purpose, the SPM 280 may have a first measuring probe for analyzing the calibration structure 225, 300 and/or the sample 205 and a second measuring probe for processing the sample 205 or the calibration structure 225, 300.

Of the SPM 280, the device 200 is the 2 only the measuring head 285 is shown. The measuring head 285 comprises in the example the 2 a holding unit 287. By means of the holding unit 287, the measuring head 285 is attached to the frame of the device 200 (in the 2 not shown). A piezo actuator 290 is attached to the holding unit 287 of the measuring head 285, which enables a movement of the free end of the piezo actuator in three spatial directions (in the 2 (not shown). Attached to the free end of the piezo actuator 290 is a probe 295 or measuring probe 295, which comprises a cantilever 294 or lever arm 294 and a measuring tip 292. The free end of the cantilever 294 of the measuring probe 295 has the measuring tip 292.

The SPM 280 can be used in the device 200 alternatively or additionally for scanning the sample 205 and/or the calibration structure 225, 300. The device 200 can employ two or more SPM 280s. The SPM 280s can be of the same type or implemented as different SPM types.

The computing unit 240 of the computer system 230 of the device 200 can have algorithms designed to determine an image of a calibration structure 225, 300 or an image of the sample 305 from measurement data of the SEM 210 and/or the SPM 280. The algorithms can be implemented in hardware, software, firmware, or a combination thereof. Furthermore, the computing unit 240 of the computer system 230 can contain algorithms that can determine the positions or the position data of the markings of the calibration structure 225, 300 from an image of a calibration structure 225, 300.

The 3 shows a schematic plan view of a calibration structure 300, which can be used to determine a concentricity error of the rotational axis 180 of the sample table 100 of the device 200. Around a central marking 330 of a substrate 370, on which a calibration structure 300 is to be produced, a first marking 310 is produced on the substrate 370 at a predetermined distance 340 from the central marking 330. The 3 The calibration structure 300 shown as an example has ten second markings 320-1 to 320-10. These are arranged at a distance 350 from the central marking 330 on a circular arc around the central marking 330 on the surface 380 of the substrate 370. However, the distance between the second markings 320-1 to 320-10 along the circular arc is not the same, so that the angles of rotation between the individual second markings have different numerical values. In the example of 3 However, the markings 310, 320-1 to 320-10 are arranged on a circular arc and thus have constant distances 340, 350 from the central marking 330. As explained above, it is also possible, however, that the distance between the first 310 and the ten second markings 320-1 to 320-10 changes depending on the angle of rotation θ.

Several possibilities for generating the markings 310, 320-1 to 320-10 and 330 of the calibration structure 300 are discussed in the context of the 2 discussed. In the exemplary calibration structure 300 of the 3 The markings 310, 320-1 to 320-10 do not extend over a complete circle, but only over a circular arc 360, which covers an angular range of approximately 0° to 300°. The marking(s) of the calibration structure 300 missing for a circle can be supplemented, for example, by extrapolating the circular arc 360.

In the exemplary calibration structure 300 of the 3 Markings 310, 320-1 to 320-10, and 330 have a square basic shape. However, markings with other basic shapes are also possible, such as round ones. Shapes of markings 310, 320-1 to 320-10, and 330 are particularly advantageous because they allow for easy determination of a reference point for the markings.

The 4 schematically illustrates an ideal calibration structure 445 in the left-hand sub-image 405. In this structure, the markings 420 are arranged on a circle 410 around a central marking 430. The center of the circle 410, designated by the reference point 440 in the sub-image 405, corresponds to the central marking 430 of the ideal calibration structure 445. In the ideal calibration structure 445, the individual markings 420 have no position deviations from a circular line 410 and are also equidistant on the circle 410. This means that the ideal calibration structure 445 has no angular errors or an error in the angle of rotation θ. Thus, in the ideal calibration structure 445, the nominal coordinates (x _n , y _n ) 410 of the individual markings 420 coincide with their real or actual positions or position data.

The right part of image 495 of the 4 schematically illustrates an example of a real calibration structure 485. To clarify the difference from the ideal calibration structure 445, the nominal coordinates 410 of the real calibration structure 485 are also shown in the partial image 495 as a circle 410 with the center point 440. The positions at which the markings 470 of the real calibration structure 485 are positioned deviate significantly from the circular shape of the nominal coordinates 410. In addition, the central marking 490 of the real calibration structure 485 does not correspond to the center 440. of the circle of the nominal coordinates 410. In the exemplary real calibration structure 485, the cause of a non-circular positioning of the individual markings 470 is a roundness error or a roundness deviation or eccentricity of the rotational axis 180 of the sample table 100 of the device 200 that lies outside the tolerance range.

The 5 schematically presents another possible error source of the rotation axis 180 of the sample table 100. As in the 4 shows the left part of the image 505 of the 5 an ideal calibration structure 545. The markings 520 of the ideal calibration structure 545 are positioned on a circle 510 around a central marking 530. The center point 540 of the circle 510 coincides in the partial image 505 with the central marking 530 of the ideal calibration structure 545. In the ideal calibration structure 545, the individual markings (x _n , y _n ) 520 lie on their nominal coordinates 510, namely the circle 510.

The right part of image 595 of the 5 illustrates the real counterpart 585 to the ideal calibration structure 545. Parallel to the 4 The partial image 595 also shows the ideal calibration structure 545. Unlike the partial image 495 of the 4 However, the real calibration structure 585 has an eccentricity error of the rotational axis 180 of the sample table 100. This means that each angle of rotation of the rotational axis 180 leads to a different axial offset (radial run-out) of the rotational axis 180. Although in the example of partial image 595 all markings 570 of the real calibration structure 585 lie on a circle, its radius is larger than the circle 510 of the ideal calibration structure 545. In addition, the central marking 590 of the real calibration structure 585 and the center point 540 of the circle of the nominal coordinates 510 of the ideal calibration structure 545 have an offset 575. The offset 575 between the center point of the ideal calibration structure 545 and the real calibration structure 585 can be determined by fitting the generated markings 570 of the real calibration structure 585 to a circle. The difference in the radius of the ideal 545 and the real calibration structure 585 corresponds to the eccentricity error of the real calibration structure 585, which is caused by an axial offset of the rotation axis 180 of the sample stage 100.

Typically, when creating a real calibration structure 225, 300 for the rotation axis 180 of the sample table as described above, the errors of the partial images 495 and 595 of the 4 and 5 and possible other error sources of the rotary axis 180. This means that by creating a real calibration structure 225, 300, the various contributions to a runout error of the rotary axis 180 can be jointly determined and corrected.

Diagram 600 of the 6 schematically shows method steps for generating a calibration structure 300 and its analysis for determining a runout error of a rotational axis 180 of a sample table 100. Furthermore, the diagram 600 explains possibilities for correcting an eccentricity of the rotational axis 180 of the sample table 100. The method begins at 605. In step 610, a substrate 370, from which a calibration structure 300 is to be produced by generating at least two markings 470, 570, is aligned with respect to the axis of the tool that is used to generate the markings 470, 570 on the substrate 370. For this purpose, the rotational axis 180 of the sample table 100 is aligned with the device 200, for example, the beam axis of the electron beam 215, when the markings 310, 320-1 to 320-10, 470, 570 are generated using the SEM 210. If the SEM 280 is used to generate markings 310, 320-1 to 320-10, 470, 570 on the substrate 370, the axis of the measuring tip 292 is aligned with the rotational axis 180 of the sample table 100. Details of this step are described in the patent specification DE 10 2020 309 638 B3 described.

A photomask, such as a defective mask, can be used as the substrate 370 on which a calibration structure 300 is produced by generating at least two markings 470, 570. Furthermore, mask blanks, including those from which masks cannot be manufactured due to irreparable defects, can serve as substrates 370 for producing a calibration structure 300. This type of substrate 370 for calibration structures 300 also has the great advantage that marking elements already applied by the manufacturer are present on their surface. An advantageous use of these marking elements is described below in the context of the discussion of 7 explained.

In step 615, a tool 210, 280 of the device 200 creates a mark 330 on the substrate 370 above the rotational axis 180 of the sample stage 100, which supports or holds the substrate 370. The mark 330 created above the rotational axis 180 is therefore also called the central mark 330. Several possibilities for creating marks 330, 430, 530 on a substrate 370 are explained above when introducing the device 200.

In the next step 620, the substrate 370 is rotated by a vector $\overrightarrow{r}$ in a plane perpendicular to the rotation axis 180. For this purpose, the translation axes in the xy plane of the sample table 100 perform a corresponding translational movement. After completion of the movement, the rotation axis 180 and the axis of the marker creation tool 210, 280 have a distance that corresponds to the magnitude of the vector $\overrightarrow{r}$ corresponds.

After completion of the translational movement of the substrate 370, the mark generation tool 210, 280 creates a first mark 310 on the substrate 370 at step 625, which has a distance $\left|\stackrel{\rightharpoonup }{r}\right|$ from the rotation axis 180 of the sample table 100.

At decision block 630, it is determined whether the number of marks 310, 320-1 to 320-10 generated on the substrate 370 reaches a predetermined maximum number. As long as this condition is not met, as explained in step 635, the substrate 370 is rotated by a predetermined angle about the rotation axis 180 of the sample stage 100 before generating another second mark 320-1 to 320-10. The method then returns to step 625 and generates another second mark 320-1 to 320-10 on the substrate 370.

However, if it is determined at decision block 630 that the maximum predefined number of marks 310, 320-1 to 320-10 has already been generated on the substrate, further mark generation ends. The process of producing a calibration structure 300 by generating a number of marks 310, 320-1 to 320-10 on the substrate 370 is complete.

In the following steps 640 to 655, the produced calibration structure 300 is analyzed to determine the errors that occur due to the rotation of the sample table 100 about the rotation axis 180 during the generation of the calibration structure 300. If the marking generation tool 210, 280 is also used to image the generated markings 310, 320-1 to 320-10 by scanning them, it is advantageous - as described in step 640 - to rotate the generated calibration structure 300 by the vector $-\overrightarrow{r}$ to be shifted so that the axis of the imaging tool 210, 280 again coincides with the rotation axis 180, which is designated in the calibration structure 300 by the central marking 330. If an imaging tool 280, 210 that is not identical to the marking generation tool 210, 280 is used to image the calibration structure 300, it is also advantageous to position the axis of the imaging tool 280, 210 used and the rotation axis 180 or the central marking 330 one above the other. If the SEM 210 of the device 200 is used as the imaging tool 280, 210, the axis of the SEM 210 is its beam axis or the beam direction of the electron beam 215. The axis of an SPM 280, on the other hand, is defined by the direction in which its measuring tip 292 points.

In step 645, the calibration structure 300 is imaged. For this purpose, the SEM 210 and/or the SPM 280 of the device 200 can scan over the calibration structure 300. Subsequently, in step 650, the positions of the markings 310, 320-1 to 320-10 on the calibration structure 300 can be determined.

In the next step 655, the differences or deviations of the measured positions 470, 570 of the generated markings 310, 320-1 to 320-10 from their nominal positions 460, 560 on the calibration structure 300 can be determined. Alternatively or additionally, the calibration structure 300 can be optically imaged. The determined deviations include all contributions that occur when rotating the rotational axis 180 of the sample stage 100. In the present application, the sum of these contributions is referred to as the runout error.

Finally, in step 660, the prerequisites for correcting the determined runout error are created. For this purpose, a table for the errors occurring when rotating the rotational axis 180 as a function of the rotation angle θ _i can be created from the determined deviations (Δx(θ _i ), Δy(θ i )). The number of markings 310, 320-1 to 320-10 present on the calibration structure 300 determines the number of table entries. Deviation values between the rotation angle ranges of adjacent markings 320-i and 320-(i+1) can be determined by interpolation. If the calibration structure 300 only has markings that extend over a circular arc, the missing part can be extrapolated to complete the table. Finally, it is also possible to construct a function (Δx(θ), Δy(θ)) from the determined deviations, including suitable interpolation and, if necessary, extrapolation. Flowchart 600 ends at 665.

The 7 illustrates an alternative and/or additional option that can be used in the event of a rotational error of the rotational axis for determining the runout error of the rotational axis 180 of the sample stage 100. Many samples have distinctive structural elements that can be used to determine a runout error. Furthermore, an important sample class, namely photomasks or mask blanks, has marking elements already applied and precisely measured by the manufacturer. The sample 700 of diagram 795 is a mask 700 or a mask blank 700. The sample 700 or mask 700 shows a schematic plan view of three marking elements 710, 730, 750, which are located in three different corners of the mask. 700 are attached. The exemplary marking elements 710, 730, 750 have the shape of rectangular angles. The reference points 720, 740, and 760 of the exemplary marking elements 710, 730, and 750 form the intersection points of the inside of the right angles. The coordinate system (u, v) associated with the photomask 700 is indicated on the sample 700. The reference points (u ₁ , v ₁ ), (u ₂ , v ₂ ), and (u ₃ , v ₃ ) in the mask coordinate system (u, v) are provided by the mask manufacturer.

At least two of the three marking elements 710, 730, 750 are scanned with the electron beam 215 of the SEM 210 and/or with the measuring tip 292 of the SPM 280 in order to determine at least two reference points 720, 740, 760 of the three marking elements 710, 730, 750. To position the marking elements 710, 730, 750 under the electron beam 215 of the SEM 210 or the measuring tip 292 of the SPM 280, the sample stage 100 exclusively performs translational movements of the sample holder 190 in the x and y directions. A rotational movement of the sample stage 100 to determine the reference points 720, 740, 760 is not performed. As soon as one of the marking elements 710, 730, 750 is placed under the electron beam 215 or the measuring tip 220, the electron beam 215 or the measuring tip 220 scans the marking element 710, 730, 750 to determine its reference point 720, 740, 760.

After scanning two marking elements 710, 730, 750, for example, the marking elements 710, 730, their reference points 720, 740 in the coordinate system (x, y) of the sample stage 100 are known: (x ₁ , y ₁ ) and (y ₁ , y ₂ ). Using the two measured reference points (x ₁ , y ₁ ) and (x ₂ , y ₂ ) as well as the reference points (u ₁ , v ₁ ) and (u ₂ , v ₂ ) provided by the mask manufacturer, a translation, rotation, and scaling of the two coordinate systems relative to each other can be determined. An affine transformation connects the measured reference points (x ₁ , y ₁ ) and (x ₂ , y ₂ ) with the reference points (u ₁ , v ₁ ) and (u ₂ , v ₂ ) provided by the mask manufacturer.

The established description of affine coordinates is done in a matrix representation using homogeneous coordinates: $$\left[\begin{array}{c}\text{u}\\ \text{v}\\ 1\end{array}\right]=\left[\begin{array}{ccc}{\text{a}}_0& {\text{a}}_1& {\text{a}}_2\\ {\text{b}}_0& {\text{b}}_1& {\text{b}}_2\\ 0& 0& 1\end{array}\right]\left[\begin{array}{c}\text{x}\\ \text{y}\\ 1\end{array}\right]$$

In the case of the alignment of the two coordinate systems to each other on the basis of two reference points, for example the reference points 720 and 740, the above general vector equation reduces to: $$\left[\begin{array}{c}\text{u}\\ \text{v}\\ 1\end{array}\right]=\left[\begin{array}{ccc}\text{s}*\text{cos }\mathrm{\alpha }& -\text{s}*\text{sin }\mathrm{\alpha }& \text{a}\\ \text{s}*\text{sin }\mathrm{\alpha }& \text{s}*\text{cos }\mathrm{\alpha }& \text{b}\\ 0& 0& 1\end{array}\right]\left[\begin{array}{c}\text{x}\\ \text{y}\\ 1\end{array}\right]$$

Here, the parameters: a, b denote a shift or offset, s a scaling, and α a rotation angle of the two coordinate systems to each other.

Inserting the measured and specified reference points 720, 740 results in a system of equations that can be solved using linear algebra methods. If the reference points 720, 740, 760 of the three marking elements 710, 730, 750 are measured for sample 700 and inserted into the general vector equation, in addition to a displacement a, b, a scaling s, a rotation α, the parameters of a shear force m and an axial reflection r can be determined.

Typically, the coordinates of an identified defect (u _D , v _D ) are specified in coordinates of the coordinate system (u, v) associated with the photomask 700. After determining the parameters of the affine transformation, these coordinates can be converted into coordinates of the sample stage 100.

After determining the relationship between the coordinate system (u, v) associated with the sample 700 and the coordinate system associated with the sample holder 190 of the sample table 100, the sample or mask 700 can be aligned such that one of the reference points 720, 740, 760 of one of the marking elements 710, 730, 750 coincides with the rotation axis 180 of the sample table 100. In the example of the 7 This is the reference point 720. When the sample 700 is rotated about the rotation axis 180, the reference points 740, 760 of the marking elements 730, 750 nominally follow a circular arc. By rotating the sample 700 at least twice by a predetermined angle of rotation 770, and by determining at least the position of one of the two reference points 720, 740 of the marking elements 730, 750 after each of the at least two rotations, the rotation errors or the concentricity error of the rotation axis 180 of the sample stage 100 can be determined from the deviations of the determined positions from the nominal positions of at least one of the reference points 720, 740 after the first and at least one second rotation. To determine the positions of the reference point(s) 720, 740 after the first or after at least a second rotation by a predetermined angle of rotation, the electron beam 215 and/or the measuring tip 292 can be used.

Analogous to the discussion above of the 6 As explained for a calibration structure 300, the runout error of the rotational axis 180 of the sample table 100 of the device 200 can be determined from the deviations of the determined position data from the nominal position data of the reference points 720, 740 of the marking elements 730, 750.

Determining the axis offset of the rotation axis, as in the 7 described, has the advantage of not requiring the creation of a calibration structure 300. Furthermore, the positions or the position data of the reference points 720, 740, 760 of the marking elements 710, 730, 750 are measured and specified by the manufacturer of the sample 700 with the greatest possible precision. The marking elements 710, 730, 750 or their reference points 720, 740, 760 thus represent a type of ideal calibration structure in which these markings are provided with the smallest possible position inaccuracies. When determining the concentricity error according to the 7 This source of error, which is caused by a variation in the positions of the generated markings 310, 320-1 to 320-10 on a calibration structure 300, can thus be largely avoided.

However, determining a runout error based on the marking elements 710, 730, 750 is disadvantageous because their spacing cannot be changed. Determining an image of the sample 700 can be complex, since the spacing between adjacent marking elements 710, 730, 750 is typically greater or much greater than the scanning range of the electron beam 215 of the SEM 210 and/or the scanning range of the measuring tip 292 of the SPM 280. Thus, imaging the marking elements 710, 730, 750 of the sample requires combining the data from multiple scanning processes into one image.

Additionally or alternatively, this can be done using the example of 7 The described rotation of a mask 700 or a mask blank 700 can also be used to determine an angular error when rotating by a predetermined angle of rotation. 7 As can be seen, an actually performed rotation by an angle θ is given by: tan θ = Δv/Δu . The determined angle of rotation θ is compared with the nominally specified angle of rotation and the determined difference or error is corrected. This means that in the context of the 6 explained table, which contains deviations (Δx(θ _i ), Δy(θ _i )) for various specified angles of rotation θ _{i of} the axis of rotation 180 of the sample table 100, the specified angles of rotation θ _i are corrected by the determined differences (Δθ _i ).

The flowchart 800 of the 8 As a first aspect of determining a runout error, FIG. 1 again presents exemplary steps of creating a calibration structure 300 on a substrate 370. The method begins at step 810.

In the first step 820, a first marking 310 is created on a surface of a substrate 370, on which the calibration structure 300 is to be produced, at a predetermined distance from the rotation axis 180 of the sample stage 100. For this purpose, the electron beam 215 of the SEM 210 in combination with an etching gas or a deposition gas and/or a measuring tip 292 of the SPM 280 of the device 200 can be used.

In step 830, the substrate 370 is rotated by a predetermined angle of rotation about the rotation axis 180, and the predetermined distance from the rotation axis 180 is changed depending on the predetermined angle of rotation. In one embodiment, the predetermined distance of the generated markings 310, 320-1 to 320-10 from the rotation axis 180 is not changed. The sample holder 190 of the sample stage 100 rotates the substrate 370 by the predetermined angle of rotation about the rotation axis 180. If necessary, the translation axes of the sample stage 100 for the x and y directions move the substrate 370 held by the sample holder 190 in the plane of the sample holder 190.

In step 840, at least one second marking 320-1 to 320-10 is generated on the substrate 370 to generate the calibration structure 300. The generation of the at least one second marking 320-1 to 320-10 is carried out analogously to the generation of the first marking 310.

The method ends at step 850.

Furthermore, the flowchart 900 of the 9 As a first embodiment of a second aspect, the method provides essential steps for determining a runout error when rotating a sample 700 about a rotation axis 180. The calibration structure 300 described in the first aspect can be used for this purpose. The method begins at step 910.

In the first step 920, at least one image of a calibration structure 300 is obtained, which has at least two markings 310, 320-1 to 320-10, the positions of which can be converted into one another by rotating the calibration structure 300 by a predetermined angle of rotation about the axis of rotation 180.

At step 930, the runout error is determined based on the at least one image. The method ends at 940.

In addition, the flowchart gives 1000 of the 10 As a second embodiment of a second aspect, essential steps of determining a runout error when rotating a sample 700 about a rotation axis 180 are provided. The calibration structure 300 described in the first aspect can be used for this purpose. Alternatively or additionally, a sample having at least one marking in the form of a structural element can be used for this purpose. Samples 700 having at least two marking elements 710, 730, 750 applied by the sample manufacturer, the position data of which are made available to the sample user, are preferred. The method begins at step 1010.

In the first step 1020, the calibration structure 300, which has at least one marking, is imaged. Then, in step 1030, the calibration structure 300 is rotated by a predetermined angle about the rotation axis 180. In step 1040, the rotated calibration structure 300 is imaged. Finally, in step 1050, the runout error is determined from the at least two images of the calibration structure 300. The method ends in step 1060.

Finally, the flowchart 1100 reproduces the 11 as a third aspect, the correction of a runout error when rotating a sample 700 about the rotation axis 180 of the sample stage 100. The method begins at step 1110.

In the next step 1120, a sample 700 is rotated by a predetermined angle of rotation about a rotation axis 180, wherein the angle of rotation is related to a predetermined orientation of the sample 700 in a plane perpendicular to the rotation axis 180.

Then, in step 1130, during rotation and/or at the end of rotation, a runout error associated with the specified rotation angle is corrected by translating the sample 700 in a plane perpendicular to the rotation axis 180. The correction can be performed using a table listing runout errors for various rotation angles of the rotation axis 180. The method finally ends at step 1140.

QUOTES CONTAINED IN THE DESCRIPTION

This list of documents submitted by the applicant was generated automatically and is included solely for the convenience of the reader. This list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• DE 10 2020 209 638 B [0029, 0070]
• DE 10 2023 205 623.2 [0029]
• DE 10 2020 309 638 B3 [0142]

Claims (21)

A method (800) for producing a calibration structure (300) for determining a runout error when rotating a sample (700) about a rotation axis (180), the method (800) comprising the steps of: a. producing (820) a first marking (310) on a surface (380) of a substrate (370) on which the calibration structure (300) is to be produced, at a predetermined distance (340) from the rotation axis (180); b. rotating (830) the substrate (370) by a predetermined angle of rotation about the rotation axis (180); and c. producing (840) a second marking (320-1 to 320-10) on the surface (380) of the substrate (370) to produce the calibration structure (300). Method (800) according to the preceding claim, further comprising: repeating steps b. and c. to generate a predetermined number of markings (310, 320-1 to 320-10) on the surface (380) of the substrate (370) of the calibration structure (300). Method (800) according to the preceding claim, wherein the predetermined number of markings (310, 320-1 to 320-10) of the calibration structure (300) comprises at least 4, preferably at least 8, more preferably at least 16, and most preferably at least 32 or 64 markings (310, 320-1 to 320-10). Procedure (800) according to Claim 2 or 3 , wherein a sum of the predetermined angles of rotation for generating the predetermined number of markings (310, 320-1 to 320-10) comprises at least 22.5°, preferably at least 45°, more preferably at least 180°, and most preferably 270° or 359°. Method (800) according to one of the preceding claims, wherein the axis of rotation (180) is oriented substantially perpendicular to the surface (380) of the substrate (370). Method (800) according to one of the preceding claims, wherein generating the calibration structure (300) comprises: aligning the rotation axis (180) of the substrate (370) of the calibration structure (300) with respect to an axis of a tool (210, 280) and generating a central marking (330) at a position of the rotation axis (180) on the surface (380) of the substrate (370). Method (800) according to the preceding claim, wherein generating the calibration structure comprises: shifting the substrate (370) by a vector $\overrightarrow{r},$ which is parallel to a plane oriented perpendicular to the axis of rotation (180) of the substrate (370), before producing the first mark (310). Method (800) according to one of the preceding claims, further comprising: imaging the at least two generated markings (310, 320-1 to 320-10) of the calibration structure (300). A method (900) for determining a runout error when rotating a sample (700) about a rotation axis (180), the method (900) comprising the steps of: a. Obtaining (920) at least one image of a calibration structure (300) having at least two markings (310, 320-1 to 320-10), the positions of which can be converted into one another by rotating the calibration structure (300) by a predetermined angle of rotation about the rotation axis (180); and b. Determining (930) the runout error based on the at least one image. The method (900) of the preceding claim, wherein obtaining the at least one image of the calibration structure (300) comprises at least one of: obtaining the at least one image from a non-volatile memory, obtaining the at least one image via a network connection, or imaging the calibration structure (300). A method (1000) for determining a runout error when rotating a sample (700) about a rotation axis (180), the method (900) comprising the steps of: a. imaging (1020) a calibration structure (300) having at least one marking (310, 320-1 to 320-10); b. rotating (1030) the calibration structure (300) by a predetermined angle of rotation; c. imaging (1040) the rotated calibration structure (300); and d. determining (1050) the runout error from the image of the calibration structure (300) and the image of the rotated calibration structure (300). Procedure (1000) according to Claim 11 , further comprising: aligning the rotation axis (180) of the sample (700) with one of the at least one marking. Procedure (1000) according to Claim 11 or 12 , wherein the calibration structure has at least two markings (710, 730, 750) with known position data. Method (900, 1000) according to one of the Claims 9 - 13 , wherein determining the runout error of the rotation axis (180) comprises: determining of position data of the at least two markings (310, 320-1 to 320-10) in the at least one image of the calibration structure (300) and/or determining position data of the at least one marking (710, 730, 750) in the at least two images of the calibration structure. Method (900, 1000) according to the preceding claim, wherein determining the runout error of the rotation axis (180) comprises: determining deviations of the position data from nominal coordinates for each of the at least two markings (310, 320-1 to 320-10) of the calibration structure (300), and/or determining deviations of the position data from nominal coordinates of the at least one marking (710, 730, 750) of the at least two images of the calibration structure. Method (900, 1000) according to one of the Claims 9 - 15 , further comprising: determining a difference between a specific angle of rotation and a nominal angle of rotation and correcting the difference to determine the runout error of the rotation axis (180). A method (1100) for correcting a runout error during rotation of a sample (700), the method (1100) comprising the steps of: a. rotating (1120) the sample (700) by a predetermined angle of rotation about a rotation axis (180), wherein the angle of rotation is related to a predetermined orientation of the sample (700) in a plane perpendicular to the rotation axis (180); and b. correcting (1130) a runout error associated with the predetermined angle of rotation during rotation and/or after rotation by translating the sample (700) in a plane perpendicular to the rotation axis (180) based on the angle of rotation. A device (200) for correcting a runout error during rotation of a sample (700), comprising: a. means (180, 255) for rotating the sample (700) by a predetermined angle of rotation about a rotation axis (180), wherein the angle of rotation is related to a predetermined orientation of the sample (700) in a plane perpendicular to the rotation axis (180); and b. means (255) for correcting a runout error associated with the predetermined angle of rotation during rotation and/or at the end of rotation by means (130, 150) for displacing the sample (700) in a plane perpendicular to the rotation axis (180), wherein the means (130, 150) for displacing the sample (700) is configured to displace the sample (700) based on the predetermined angle of rotation. Device (200) according to Claim 18 , further comprising: means (210, 280) for imaging the sample (700) while rotating the sample (700). Device (200) according to Claim 18 or 19 , wherein the device (200) is arranged to carry out the method steps of Claims 1 until 17 to execute. Computer program containing instructions for carrying out the method steps of any of the Claims 1 until 17 includes.