The present invention relates to a method for determining a geometry of a meas- uring tip for a scanning probe microscope, a test structure, a test device and a de- vice for analysing and/or processing a sample.

The content of the priority application DE 10 2020 118 150.7 is incorporated by reference in its entirety.

Microlithography is used for producing microstructured components, such as, for example, integrated circuits. The microlithography process is performed using a lithography apparatus, which has an illumination system and a projection sys- tem. The image of a lithography mask (reticle) illuminated by means of the illu- mination system is in this case projected by means of the projection system onto a substrate, for example a silicon wafer, which is coated with a light-sensitive layer (photoresist) and arranged in the image plane of the projection system, in order to transfer the mask structure to the light-sensitive coating of the sub- strate.

To obtain smaller structure sizes, and hence an increase in the integration den- sity of the microstructured components, use is increasingly made of light with very short wavelengths, which is referred to as deep ultraviolet (DUV) or extreme ultraviolet (EUV), for example. DUV has a wavelength of 193 nm, for example, and EUV has a wavelength of 13.5 nm, for example. Here, the lithography masks themselves have structure sizes ranging from 5 to 100 nm. The production of such lithography masks is very complicated and hence costly, in particular since the lithography masks have to be defect free because otherwise it is not possible to ensure that a structure generated by the lithography mask has the desired function. For this reason, lithography masks are verified, for example, i.e., the defect-free property of the lithography mask is tested. In the process, defects are recognized and localized, facilitating a targeted repair of the defects. Typical de- fects include the lack of envisaged structures, for example because an etching process was not carried out successfully, or else the presence of non-envisaged structures, for example because an etching process proceeded too quickly or de- veloped its effect at a wrong position. These effects can be remedied by targeted etching of excess material or targeted deposition of additional material at the ap- propriate positions; by way of example, this is possible in a very targeted manner by means of electron beam-induced processes (FEBIP, "focussed electron beam in- duced processing").

The advance of the process is monitored continuously during the repair process, preferably by means of a scanning electron microscope. The electron microscope offers a spatial resolution suitable to this end, which lies in the region of a few nanometres. However, it is disadvantageous that the surface to be examined is only representable on the basis of contrasts on account of different interactions of the electron beam with the surface atoms. Therefore, the image of an electron mi- croscope typically contains no information about the height of the surface.

Scanning probe microscopy (SPM) is an imaging process which allows height in- formation about the surface to be examined to be ascertained. To this end, use is made of a measuring probe which directly interacts with the surface and which is raster scanned over the surface. In the case of atomic force microscopy (AFM), the interaction can be based on direct contact, on a van der Waals interaction or on further physical interactions and mixtures thereof. By way of example, the inter- action is kept constant for each raster position of the measuring probe by virtue of a height of the measuring probe above the surface being set by means of a mi- cro-actuator. The measuring probe is also referred to as a tip. The geometry of the tip has a significant influence on the interaction between the tip and the sur- face, particularly in the case of sharp structures. The result of such a measure- ment approximately corresponds to a convolution of the tip geometry with the surface geometry. Therefore, it is advantageous to accurately know the tip geom- etry in order to be able to correctly interpret the measurement result.

The following documents consider how to take account of the influence of the ge- ometry or form of a measuring tip of an SPM on SPM images of a sample surface : V. Bykov et al.· “Test structure for SPM tip shape deconvolution”, Appl. Phys. A 66, p. 499-502 (1998); G. Reiss et al.· “Scanning tunneling microscopy on rough surfaces: Deconvolution of constant current images”, Appl. Phys. Lett. 57 (9), 27 August 1990, p. 867-869; Y. Martin and H.K. Wickramasinghe: “Method of imag- ing sidewalls by atomic force microscopy”, Appl. Phys. Lett. 64 (19), 9 May 1984, p. 2498-2500; L. Martinez et al.· “Aspect-ratio and lateral-resolution enhance- ment in force microscopy by attaching nanoclusters generated by an ion cluster source at the end of a silicon tip”, Rev. Sci. Instrum. 82, (2011), p. 02370- 1 - 023710-7; X. Qian et al.· “Image simulation and surface reconstruction of under- cut features in atomic force microscopy”, SPIE Proc. Vol. 6518, (2007), p. 1-12; L. Udpa et al.· “Deconvolution of atomic force microscopy data for cellular and mo- lecular imaging”, IEEE, Sig. Proc. Mag. 23, 73 (2006); Ch. Wong et al.· “Tip dila- tion and AEM capabilities in the characterization of nanoparticles”, J. O. Min. 59, 12 (2007); J.S. Villarrubbia ^: “Algorithms for scanned particle microscope image simulation, surface reconstruction, and tip estimation”, J. Res. Natl. Inst. Stand. Technol. 102, p. 425-454 (1997); X. Qian and J.S. Villarrubia: “General three-di- mensional image simulation and surface reconstruction in scanning probe mi- croscopy using a dexel representation”, Ultramicroscopy 1008 (2007), p. 29-42.

By way of example, using test structures whose geometry is known and which are measured by the measuring tip allows the geometry of the tip to be deduced by "deconvolving" the measuring signal with the known test structure geometry.

The tip geometry subsequently known can be used to "deconvolve" the measure- ment signal of an unknown structure and consequently draw conclusions about the actual geometry thereof. However, a measuring tip changes during its use on account of the constant interaction with the surface, which is why the tip geome- try should be checked at regular intervals. This may be very time-consuming, particularly if the measuring tip needs to be displaced to a separate test struc- ture or even needs to be disassembled from the measurement device to this end.

It is therefore desirable to have an option for characterizing the geometry of the measuring tip which is exact but nevertheless does not cause much outlay, in particular requires little time.

DE 10 2016 223 659 A1 describes a method for providing a measuring tip for a scanning probe microscope. At least a first and at least a second measuring tip are provided in a first step, wherein the first and the second measuring tip are arranged on a common measuring tip carrier in such a way that only the first measuring tip interacts with a sample to be examined. The first measuring tip is irreversibly processed in a second step such that, following the irreversible pro- cessing, only the second measuring tip interacts with the sample to be examined. DE 10 2018 221 304 A1 describes a device for in situ determination of a process resolution of a particle beam-induced processing process of a photolithographic element. The latter comprises a processing unit embodied to generate at least two reference structures on the element within a process environment, said reference structures differing in at least one dimension, and an analysis unit embodied to determine the dimensions of the at least two reference structures in the process environment in order to determine the process resolution of the particle beam-in- duced processing process therefrom.

DE 10 2017 211 957 A1 describes a method for examining a measuring tip of a scanning probe microscope, wherein a test structure is generated and the meas- uring tip is examined with the aid of the generated test structure.

Against this background, it is an object of the present invention to improve the use of a measuring tip in processes for analysing or processing surfaces, in partic- ular microstructured surfaces.

According to a first aspect, a method for determining a geometry of a measuring tip for a scanning probe microscope is proposed. In a first step a), at least one test structure is generated, the latter having elevations alternating with depressions in a first direction, wherein the elevations and depressions are aligned parallel to one another in a second direction perpendicular to the first direction. In a second step b), the test structure is scanned by the measuring tip in the first direction and in a third direction that is perpendicular to the first and second direction, for the purposes of ascertaining a profile of the test structure. In a third step c), the geometry of the measuring tip is ascertained on the basis of the ascertained pro- file.

In particular, the test structure is generated with a target structure size. The test structure has at least one depression.

This method is advantageous in that the geometry of the measuring tip can be ascertained quickly and easily using a comparatively simple test structure. On account of the parallelism of the structures, it is possible to reliably avoid the measuring tip (partially) missing the test structure, which may be the case, for instance, where there are column-like elevations or depressions, and which may lead to measurement errors. In particular, this can be implemented in situ, with the test structure being generated specifically to this end on a sample surface to be examined, for example. The test structure is able to be generated or produced accurately on account of its simple geometry. Moreover, relevant dimensions of the test structure can be easily ascertained and hence checked by means of com- plementary measurement means such as an electron microscope, for example. Furthermore, the geometry of the measuring tip can be ascertained in targeted fashion at different flanks of the measuring tip by means of the proposed test structure.

A scanning probe microscope can be operated in different modes of operation. In addition to being used as an imaging measuring instrument which renders a three-dimensional structure of a surface able to be captured, the measuring tip can also be used as a micro-manipulator. What the measuring modes of operation have in common is that the sample surface to be captured is raster scanned with the measuring tip. A height value, i.e., a value of a z-coordinate, is ascertained for each measurement point which is determined by a tuple of (x, y) -coordinates for example. The horizontal resolution and also vertical resolution of a scanning probe microscope are in the region below one nanometre. By way of example, the scanning probe microscope is embodied as an atomic force microscope.

By way of example, the measuring tip is arranged at a free end of a cantilever. The end of the cantilever opposite to the free end is held by a probe holder or measuring head, which is displaceable in all three spatial directions by means of a piezo actuator system, for example. Additionally, provision can be made for the probe holder to be mounted in rotatable and/or tiltable fashion. The basic resolu- tion of the scanning probe microscope is determined by the piezo actuator system, for example, since the latter is responsible for exactly approaching the measure- ment points arranged in a raster. Overall, the cantilever with the measuring tip arranged thereon can be referred to as a measuring probe. By way of example, the cantilever consists of monocrystalline silicon; however, other materials and/or material combinations may also be provided depending on the application. Different types of scanning probe microscopes are known, for example scanning tunnelling microscopes and atomic force microscopes. A voltage is applied be- tween the measuring tip and the sample surface in the case of scanning tunnel- ling microscopes and the current that flows without direct contact between tip and surface (the "tunnelling current") is measured. The current is an indicator for the distance between the measuring tip and the sample surface. The measur- ing tip directly interacts with the sample surface at an atomic level in the case of atomic force microscopes. This will be explained in slightly more detail below.

The measuring tip interacts with the sample surface for the purposes of captur- ing the height of the sample surface at a respective measurement point. By way of example, the measuring tip is brought into contact with the sample surface. To this end, the measuring probe is made to approach the sample surface from above. Contact is established as soon as bending of the cantilever is determined; by way of example, this is ascertained optically by way of the deflection of a laser beam reflected from the upper side of the cantilever. The measuring tip is subse- quently made to maintain contact with the surface and moved (raster scanned) over the surface. The bend of the cantilever is a measure for the height of the sur- face of the respective raster point. Thus, a height of the probe holder for example is kept constant in this mode. As an alternative thereto, it is possible to keep the bend of the cantilever constant by virtue of controlling the height of the probe holder by means of an appropriate control loop. Then the displacement of the probe holder in the z-direction corresponds to the height of the sample surface. Further measurement modes such as a non-contact mode or an intermittent mode, which differ in terms of the type of interaction of the measuring tip with the surface in particular, are possible in addition to the contact mode, where the measuring tip wears out comparatively quickly. By way of example, in these modes, the cantilever is brought into forced oscillation at a given frequency, which is preferably close to a resonant frequency of the cantilever. The oscihation is excited in such a way that the free end of the cantilever with the measuring tip oscillates in the z-direction. In this case, the resonant frequency depends in par- ticular on the properties of the cantilever and an interaction present between the measuring tip and the sample surface. By way of example, the resonant fre- quency and hence the amplitude of the forced oscillation at the specified fre- quency change as soon as there is a change in the interaction. By way of example, the z-direction of the probe holder is updated by a closed control loop in such a way that the oscillation is maintained, as a result of which the displacement of the probe holder in the z-direction corresponds to the height of the sample sur- face. These measurement methods described in exemplary fashion are able to be extended and/or modified in various ways. Further, various other measurement methods are possible.

The measuring tip itself has a pyramid-like or needle-like form, for example. De- pending on the application, a particularly fine measuring tip is advantageous, for example if very fine structures with a high aspect ratio should be measured using the measuring tip. The geometry of the measuring tip has a significant influence on the profile of the sample surface captured by means of the measuring tip, and so it is advantageous to know the geometry as accurately as possible.

There may be erroneous or incomplete measurement results depending on the ap- pearance of the geometry of a measuring tip. Two examples in this respect: A measuring tip with a tip diameter of 60 nm, for example, cannot capture a de- pression, for example a trench or a hole in the sample surface, with a width or di- ameter of only 30 nm since the measuring tip cannot dip into the depression. Likewise, a measuring tip with a length of 50 nm cannot completely capture a de- pression with a height of 100 nm.

The geometry of the measuring tip can be characterized by various parameters. By way of example, a diameter and a length can completely characterise a rod- shaped measuring tip. If the measuring tip has a different basic shape, for exam- ple a cone-like or pyramid-like shape, an opening angle of the cone or of the pyra- mid can also be a useful parameter. However, a measuring tip can also have dif- ferent geometries at different flanks, and so the geometry of the measuring tip is preferably determined in relation to a respective flank. Such geometry-character- izing parameters can be ascertained by scanning the test structure and subse- quently analysing the profile captured in the process.

By way of example, the test structure is generated on a surface of a sample to be examined. The surface on which the test structure is generated of the sample to be examined is preferably flat and smooth. In particular, "flat" is understood to mean that the surface lies in the plane spanned by the first and the second direc- tion, at least in the region of the test structure. In particular, "smooth" is under- stood to mean that the surface prior to the generation of the test structure has no substantial unevenness or no structures deviating from this plane. By way of ex- ample, "substantial unevenness" should be understood to mean unevenness of more than 10% of the structure size of the test structure. By way of example, if the test structure has an elevation with a height of 20 nm and a width of 10 nm, the boundary from insubstantial to substantial would be at 2 nm height and 1 nm width. Depending on application, the understanding of a person skilled in the art as to what is substantial or insubstantial may deviate from this example.

The test structure is generated with a target structure size. That is to say, the generated test structure is generated according to predetermined target values which the test structure should exhibit. In the present case, the term "structure size" should be understood to mean geometric parameters in particular, which describe a shape of the test structure. In particular, these comprise a width of the at least one depression, i.e., a distance between the flanks of two adjacent eleva- tions, a width of an elevation, a level difference between a depression and an ele- vation (can also be referred to as "height" or as "depth") and/or a flank angle of a flank between a depression and an elevation, wherein the flank angle is an angle between a flank and a plane spanned by the first direction and the second direc- tion.

Depending on the process used to generate the test structure, the generated test structure can have deviations from the predetermined structure size. In particu- lar, the deviations are random and do not relate to the test structure overall but only a certain aspect of the test structure in each case. In particular, these are not systematic deviations. The better the employed process for generating the test structure is known and the more stably the latter is able to be performed, the more reliably the test structure can be produced with the target structure size. This contributes to reliable determination of the geometry of the measuring tip since the target structure size is preferably included as a parameter when deter- mining the geometry on the basis of the ascertained profile. Therefore, an incor- rect assumption about the structure size of the test structure is reflected, where applicable, in an incorrect determination of the geometry. The at least one test structure has elevations alternating with depressions in a first direction, which are aligned parallel to one another in a second direction perpendicular to the first direction. By way of example, the elevations are struc- tures that are elevated in relation to a lower point in a depression and, by way of example, the depressions are recesses in relation to an upper point of an eleva- tion. That is to say, the terms elevation and depression should be understood rel- ative to one another. By way of example, elevations and depressions can be repre- sented in the form of at least one trench in a smooth surface. Then the surface lo- cated adjacent to the depression (the trench) corresponds to an elevation. Con- versely, in relation to a web representing an elevation in respect of the surface, the surface adjoining the web can form a depression. The elevations and depres- sions preferably extend in the direction of a surface normal in relation to the sur- face carrying the test structure. One could also say that the flanks of the test structure run perpendicular to the surface. However, deviations therefrom are likewise possible.

Elevations alternating with depressions means that, in particular, if a cut is made along the first direction of the test structure elevations and depressions are found present in alternating fashion. Here, the various elevations and depres- sions can have different embodiments, i.e., for example, have a variation in the width, height, flank angle and the like. Preferably, the test structure has a peri- odic configuration, at least in sections, such that a plurality of elevation/depres- sion pairs following one another in the first direction each have the same embodi- ment, i.e., have an identical geometry. Such a regular test structure allows the profile to be ascertained with significant statistics at measurement points for the purposes of determining the geometry of the measuring tip.

In embodiments the test structure comprises exactly one depression that is flanked by two elevations. One could also say that this test structure is embodied as a trench. Since the width of the trench is known as a result of the target struc- ture size, this test structure allows determination of a diameter of the measuring tip along a line extending (substantially) perpendicular to the trench flanks or along a line in a plane parallel to the plane spanned by the first direction and the second direction. If the measuring tip dips into the trench, the diameter is smaller than the width, and if the measuring tip does not dip in, the diameter is greater than the width. If the measuring tip only dips a short distance into the trench but does not reach the base of the latter, the measuring tip has a diameter along its profile that is greater than the width; however, a front end of the meas- uring tip has a smaller diameter. In the case of an arrangement of a plurality of trenches parallel to one another, this applies accordingly to each individual trench. The diameter of the measuring tip is of particular interest since the latter determines whether a certain structure size is even able to be captured or re- solved with the measuring tip. If the measuring tip does not dip into the trench or only partly dips into the trench, then two points of the test structure, for exam- ple, are simultaneously contacted or scanned by the measuring tip at the deepest point of the ascertained profile (i.e., the data point of the profile corresponding to the deepest dip of the measuring tip). By way of example, these two points are op- posite points of a respective edge that marks the transition from the respective elevation into a flank to the depression. By way of example, the spacing of these points corresponds to the width of the depression at the relative position of the test structure. The described situation in which the measuring tip simultane- ously contacts two points of the test structure is reached, in particular, if a scan increment of the AFM (resolution of the raster points) is small in relation to the tip diameter, i.e., for example, 1:5, 1:10, 1:20 or even smaller, and/or if use is made of a measuring tip that is flexible in the scan direction (i.e., the latter possi- bly bends when dipping into the respective trench such that it can simultane- ously contact two points of the test structure).

In particular, the elevations and depressions being aligned parallel to one an- other in a second direction perpendicular to the first direction is understood to mean that, for example, a tangent placed against an edge between an elevation and the adjacent depression extends in the second direction. Hence, the eleva- tions and depressions can have a curved profile, in the sense that the tangents of two spaced apart points on the edge need not be parallel to one another.

Preferably in step b) the test structure is scanned by the measuring tip in the first direction and in a third direction that is perpendicular to the first and sec- ond direction, for the purposes of ascertaining a profile of the test structure. As described above, in step b), scanning is implemented in such a way that the probe holder with the measuring probe raster scans the test structure. A height value is ascertained at each raster point of the raster. Here, the height value de- scribes a distance of the point on the test structure corresponding to the raster point from a reference level given for example by a surface of a sample carrying the test structure, said distance being measured in a third direction orthogonal to the first direction and the second direction. Below, the third direction is also re- ferred to as z- direction.

Scanning the test structure with the measuring tip is preferably implemented in line-like fashion, for example along the first direction but directions deviating therefrom are also possible. By way of example, a line comprises 1000 raster points in this case and extends over a distance of 1 μm. Hence, the resolution of the raster points along the first direction is 1 nm. A respective line is preferably scanned in the forward and backward direction, as a result of which each raster point is captured twice. In the process, a height value is assigned to each raster point. Once a first line has been captured, the measuring tip is offset perpendicu- lar to the first line by one raster point and a second line, which is parallel to the first line, is scanned. In this way, for example, a 1 μm ² area of the test structure is scanned with a spatial resolution of 1 nm ² . Consequently, the profile ascer- tained thus comprises 10 ⁶ (one million) measurement points. Depending on the application, the resolution in the different directions can be varied. By way of ex- ample, the deflection along the first and/or second direction can be up to 10 μm or up to 100 μm. Furthermore, in some embodiments a profile may only comprise a few lines, for example ten lines, down to only a single line. Fewer lines are cap- tured in a shorter period of time, which may be advantageous. On the other hand, fewer measurement points are then present in order to determine the geometry of the measuring tip. Moreover, a varying resolution might be used, wherein a higher resolution is used in a region of interest, in which, for example, a sharp edge of the test structure is arranged, than in other regions where the test struc- ture is flat.

The ascertained profile can be evaluated using statistical processes. In particu- lar, the ascertained and evaluated profile is used in step c) of the method accord- ing to the first aspect. By way of example, it is possible to ascertain a mean value in respect of a width of depressions as per the ascertained profile and/or in re- spect of a penetration depth of the measuring tip into the depressions as per the ascertained profile. The geometry of the measuring tip can be ascertained with a greater reliability on the basis of such a statistical evaluation since individual, random and local variations in the structure size of the test structure are "aver- aged out" in this case.

In the third step c), the geometry of the measuring tip is ascertained on the basis of the ascertained profile. It should be observed that the profile obtained by scan- ning does not reproduce the height profile of the test structure but corresponds to a convolution of the test structure geometry with the measuring tip geometry. However, it is possible to determine the test structure geometry or else the meas- uring tip geometry from the profile by deconvolving the profile. To this end, the respective other geometry should be known to the best possible extent. In the present case, the test structure is generated according to precise specifications and so the geometry thereof is known. Therefore, the geometry of the measuring tip can be determined from the ascertained profile.

In particular, the geometry of the measuring tip is ascertained or determined on the basis of:

(i) the ascertained profile,

(ii) the ascertained profile and the target structure size,

(iii) the ascertained profile and an ascertained actual structure size, or

(iv) the ascertained profile, the target structure size and an ascertained actual structure size.

Steps b) and c) can be carried out repeatedly, for example before and after a structure to be examined has been scanned by the measuring tip (step a) of the method according to the second aspect). In this way, it is possible to identify changes in the geometry of the measuring tip, which have an effect on a measure- ment result, in a timely fashion and take these into account accordingly.

The structure size of the test structure is preferably ascertained in a test or cali- bration step (in particular following step a) and before the step c) of the method according to the first aspect and/or before step a) of the method according to the second aspect). That is to say that the test structure is generated with certain process parameters and subsequently measured using one or more characteriza- tion processes.

By way of example, this can be carried out once for a number of samples to be an- alysed/processed in step a) of the method according to the second aspect, wherein the number is greater than 1, 5 or 10. The actual structure size, ascertained thus, of the test structure used for the calibration step is then preferably included in the method according to the first aspect as target structure size. Provided, as de- scribed above, a statistical distribution of the actual structure size over the test structure is ascertained when ascertaining the actual structure size, the target structure size can be determined by a value and an associated confidence inter- val, for example. Here, the confidence interval is typically a characteristic of an extent of the statistical distribution of the ascertained actual structure size. In particular, the confidence interval corresponds to a statistical moment of the dis- tribution. By way of example, the target width (an example of the target struc- ture size) of the depressions is determined as 15±0.5 nm, where 15 nm is a mean value of all actual widths (an example of the actual structure size) ascertained on the basis of the test structure used for the calibration step and 0.5 nm is the mean deviation of all ascertained actual widths from the mean value.

Alternatively, the above characterization is carried out before the analysis/pro- cessing of each sample. The actual structure size ascertained thus is then prefer- ably included in the method according to the first aspect as actual structure size.

The characterization processes comprise capturing an electron microscope image, an electron microscope image being able to capture a fracture edge of the test structure in particular, capturing a profile of the test structure with a calibrated measuring tip, the geometry of which being precisely known, and/or capturing a transmission electron microscope image (TEM image) of a lamella (TEM lamella) detached from the generated test structure, wherein the lamella is, for example, "cut out" of the surface with a focussed ion beam, received by a micromanipulator and displaced into a suitable position for capturing the TEM image. According to one embodiment of the method, an electron microscope image of the at least one test structure is captured and a structure size of the at least one test structure is ascertained from the captured electron microscope image. The expla- nations given above in respect of time and repetition of these steps apply accord- ingly. In particular, since these steps are easy to carry out, they can be carried out during each step b) of the method according to the second aspect (i.e., for each sample), to be precise between steps a) and c) of the method according to the first aspect. In step c), the geometry of the measuring tip is determined on the basis of the ascertained profile and the ascertained structure size.

In particular, the structure size ascertained here is the actual structure size of the test structure.

This embodiment is advantageous in that a structure size of the generated test structure is checked using a reliable process. The electron microscope image of the test structure allows accurate determination of dimensions of the test struc- ture in the plane spanned by the first direction and the second direction. Hence, the structure size of the test structure is accurately known in at least two dimen- sions. The structure parameters ascertained from the electron microscope image are referred to as length and width, for example. Here, a distinction should be made between an overall length/overall width which describes the test structure overall and a length/width of individual elevations or depressions which are a constituent part of the test structure. In particular, the structure size relates to the length, width and/or height of individual elevations or depressions.

In particular, the electron microscope image is captured in situ directly after the generation of the test structure or else after the scanning of the test structure with the measuring tip. The electron microscope image can be captured with an oblique angle of incidence of the electron beam. Using this, it is also possible to derive height information in respect of the elevations and depressions from the electron microscope image. By way of example, the electron beam in this case is radiated in at an angle of 40°-60°, preferably 45°-55°, further preferably 49°-51°, with respect to the plane spanned by the first direction and the second direction. In embodiments, the generated test structure is scanned using a calibrated meas- uring tip in order to ascertain the structure size of the test structure. The geome- try of the calibrated measuring tip is precisely known in this case. This is prefer- ably implemented in situ, directly after generating the test structure. Therefore, the actual structure size of the test structure can be deduced from the ascer- tained profile of the test structure in this case.

In embodiments, a TEM lamella of the generated test structure is detached and a TEM image of the TEM lamella is captured in order to ascertain the structure size of the test structure. This is preferably implemented in situ, directly after generating the test structure.

According to a further embodiment of the method at least two test structures are generated in step a), said test structures having different aspect ratios in relation to the elevations or depressions, with the aspect ratio specifying a height relative to a width of the elevation or depression. The at least two test structures are scanned by the measuring tip in step b) for the purposes of ascertaining a respec- tive profile.

Consequently, the two test structures have a different target structure sizes in particular.

It is possible to determine the geometry of the measuring tip very accurately by varying the aspect ratio. By way of example, a first test structure has rectangular elevations/depressions with a height of 50 nm and a width of 25 nm. Therefore, the aspect ratio is 2:1. The second test structure has rectangular elevations/de- pressions with a height of 50 nm and a width of 10 nm. Therefore, the aspect ra- tio is 5:1. By way of example, should scanning these two test structures yield that the profile of the first test structure has a height of 50 nm between an upper point of an elevation and a lower point of a depression then the measuring tip has a front section with a length of more than 50 nm and a diameter (along the scan- ning direction, which is located along the first direction in this case) of less than 25 nm. Further geometry details about the front section of the measuring tip may be able to be ascertained from the steepness of the flanks of the profile at the transition from elevation to depression, and vice versa. By contrast, the profile of the second test structure has only 10 nm as greatest height between an upper point of an elevation and a low point of a depression. From this, it is possible to determine that the measuring tip cannot dip into the depressions completely and also does not reach the base of the depression. By way of example, it is also possi- ble to derive that the front ten nanometres of the measuring tip have a diameter of less than ten nanometres.

It is possible to determine the geometry of the measuring tip very accurately by the combination of a multiplicity of different test structures with very different aspect ratios.

The aspect ratio is a measure as to how "fine" or "thin" a structure is. However, the aspect ratio on its own does not define the geometry of a test structure; in- stead, one of the two dimensions, i.e., the height or the width, must also be de- fined.

According to a further embodiment of the method, the at least two test structures are scanned sequentially in succession with an increasing or decreasing aspect ratio. The scanning is terminated if a deepest point of a depression is ascertained to have been reached on the basis of the ascertained profiles in the case of a se- quence with a decreasing aspect ratio or a deepest point of a depression is ascer- tained to have no longer been reached on the basis of the ascertained profiles in the case of a sequence with an increasing aspect ratio.

This sequential procedure is advantageous in order to determine, in a targeted manner, a limit that is still able to be measured or captured without errors by the measuring tip. Here, the sequence with the decreasing aspect ratio can be advan- tageous in that the front section of the measuring tip is characterized very accu- rately or in very detailed fashion. The sequence with increasing aspect ratio can be advantageous in that basic properties of the measuring tip can be ascertained in the case of a smaller aspect ratio that is resolved without problems by the measuring tip and this also allows checking of the successful generation of the test structures. As an alternative to a sequentially increasing or decreasing aspect ratio of the test structures to be scanned in succession, an alternating scheme is also possi- ble, wherein a row comprising every second test structure to be scanned has a se- quentially increasing aspect ratio or sequentially decreasing aspect ratio and a row with the test structures to be scanned lying therebetween likewise has a se- quentially increasing aspect ratio or sequentially decreasing aspect ratio.

According to a further embodiment of the method, at least one first test structure and one second test structure having elevations alternating with depressions are respectively generated in a first direction in step a), the first direction of the first test structure including an angle not equal to 0°, in particular an angle ranging from 45° to 90°, with the first direction of the second test structure. The at least two test structures are scanned by the measuring tip in step b) for the purposes of ascertaining a respective profile.

This embodiment is advantageous in that the geometry of the measuring tip can be determined even more accurately. In particular, a flank geometry of different flanks of the measuring tip can be determined precisely.

By way of example, the first test structure is scanned along its first direction, which is referred to as x-direction in this example. Thus, the probe holder moves the measuring tip back and forth along the x-direction. The trajectory of the measuring tip in the x-direction therefore strikes an edge between elevation and depression in perpendicular fashion. In this case, the geometry of the front flank or the back flank of the measuring tip is decisive for the ascertained profile. The second test structure is also scanned along its first direction, which is however perpendicular to the first direction of the first structure and which is therefore referred to as y-direction. Thus, the probe holder moves the measuring tip back and forth along the y-direction. In this case, too, the measuring tip strikes an edge between elevation and depression in perpendicular fashion. If the measur- ing tip or the probe holder has the same alignment as previously, it is now the lateral flanks of the measuring tip that are decisive for the ascertained profile.

By way of example, further or other flank regions can also be determined more accurately by virtue of the probe holder being rotated through a certain angle such that a different flank region is decisive for the ascertained profile when the measuring tip approaches an edge.

According to a further embodiment of the method, steps a) and b) are imple- mented in the same vacuum chamber and/or without the vacuum being broken.

Thus, the test structure is generated in situ, in particular, and subsequently scanned without the vacuum being broken. Complicated channelling of a test structure and/or of the measuring tip into a process environment is consequently dispensed with. In particular, a process environment has an atmosphere with very accurately controlled physical-chemical properties such as a pressure, tem- perature and a composition, in particular of the partial pressure of different spe- cies present in the atmosphere.

According to a further embodiment of the method, step a) comprises a generation of the at least one test structure with the aid of a focussed particle beam, in par- ticular with an electron beam, and at least one precursor gas or etching gas.

Structures can be generated with a very high accuracy, in particular a very high spatial resolution, by means of particle beam-induced processes. Thus, it is possi- ble to generate test structures with very small structure sizes, which are in the atomic range. By way of example, the test structures generated thus can have el- evations with a width of 1 nm alternating with depressions with a width of 1 nm and an aspect ratio of 10: 1 or more. This is particularly advantageously possible by means of election beam-induced processes (EBIP).

In particular, alkyl compounds of main group elements, metals or transition ele- ments can be considered as precursor gases suitable for the deposition or for growing of elevated structures. Examples of this include cyclopentadienyl(trime- thyl)platinum (CpPtMe ₃ Me = CH ₄ ), methylcyclopentadienyl(trimethyl)platinum (MeCpPtMe ₃ ), tetramethyltin (SnMe ₄ ), trimethylgallium (GaMe ₃ ), ferrocene (Cp ₂ Fe), bisarylchromium (Ar ₂ Cr), and/or carbonyl compounds of main group ele- ments, metals or transition elements, such as, e.g., chromium hexacarbonyl (Cr(CO) ₆ ), molybdenum hexacarbonyl (Mo(CO) ₆ ), tungsten hexacarbonyl (W(CO) ₆ ), dicobalt octacarbonyl (Co ₂ (CO) ₈ ), triruthenium dodecacarbonyl (Ru ₃ (CO) ₁₂ ), iron pentacarbonyl (Fe(CO) ₅ ), and/or alkoxide compounds of main group elements, metals or transition elements, such as, e.g., tetraethoxysilane (Si(OC ₂ H ₅ ) ₄ ), tetraisopropoxytitanium (Ti(OC ₃ H ₇ ) ₄ ), and/or halide compounds of main group elements, metals or transition elements, such as, e.g., tungsten hex- afluoride (WF6), tungsten hexachloride (WCl6), titanium tetrachloride (TiCl ₄ ), bo- ron trifluoride (BCl ₃ ), silicon tetrachloride (SiCl ₄ ), and/or complexes with main group elements, metals or transition elements, such as, e.g., copper bis(hex- afluoroacetylacetonate) (Cu(C ₅ F ₆ HO ₂ ) ₂ ), dimethylgold trifluoroacetylacetonate (Me ₂ Au(C ₅ F ₃ H ₄ O ₂ )), and/or organic compounds such as carbon monoxide (CO), carbon dioxide (CO ₂ ), aliphatic and/or aromatic hydrocarbons, and more of the same. By way of example, the etching gas may comprise: xenon difluoride (XeF ₂ ), xenon dichloride (XeCl ₂ ), xenon tetrachloride (XeCl ₄ ), steam (H ₂ O), heavy water (D ₂ O), oxygen (O ₂ ), ozone (O ₃ ), ammonia (NH ₃ ), nitrosyl chloride (NOCl) and/or one of the following halide compounds: XNO, XONO ₂ , X ₂ O, XO ₂ , X ₂ O ₂ , X ₂ O ₄ , X ₂ O ₆ , where X is a halide. Further etching gases for etching one or more of the depos- ited test structures are specified in the applicant’s US patent application having the number 13/0103281. Further additional gasses that can be used when generating the test structure comprise, e.g., oxidizing gases such as hydrogen peroxide (H ₂ O ₂ ), dinitrogen oxide (N ₂ O), nitrogen oxide (NO), nitrogen dioxide (NO ₂ ), nitric acid (HNO3) and fur- ther oxygen-containing gases, and/or halides such as chlorine (Cl ₂ ), hydrogen chloride (HCl), hydrogen fluoride (HF), iodine (I ₂ ), hydrogen iodide (HI), bromine (Br ₂ ), hydrogen bromine (HBr), phosphorus trichloride (PCl3), phosphorus pen- tachloride (PCl ₅ ), phosphorus trifluoride (PF ₃ ) and further halogen-containing gases, and/or reducing gases, such as hydrogen (H ₂ ), ammonia (NH ₃ ), methane ( CH ₄ ) and further hydrogen-containing gases. By way of example, these addi- tional gases can find more use for etching processes, as buffer gases, passivation means and the like. A depth of depressions etched into the respective substrate may not be easily de- terminable, especially in the case of etching processes. This applies even more, the greater the aspect ratio of the generated test structure is. To generate the test structure with the target structure size, in particular with a depth of the de- pression that is known in advance, it is possible to use so-called end pointing on the basis of a layer structure of the substrate known in advance. By way of exam- ple, it is known that the substrate comprises a first layer made of a first material with a certain layer thickness. A second layer made of a second material follows below the first layer. Depending on which first and second materials are present, it is possible to select and carry out an etching process which is selective for the first material, i.e., the second material is not etched or is etched with a signifi- cantly reduced etching rate in relation to the first material, for example 1 :0 or better. The first layer being completely removed and the second material being reached can be ascertained for example on the basis of a material contrast, which is visible in a secondary electron signal, between the first material and the sec- ond material. This can ensure that a depth of the depressions corresponds to the layer thickness of the first material.

According to a further embodiment of the method, the at least one test structure has a line-like structure with a comb-like cross section in a plan view, wherein the comb -like cross section in particular comprises a substantially rectangular profile.

Ideally, the test structure comprises a rectangular profile. Since the particle beam, for example the electron beam, typically has a Gaussian beam profile, the edges of a constructed structure are not sharp but rounded at an atomic level. Such structures with rounded corners are also understood to be "substantially rectangular". What structures should still be considered "substantially rectangu- lar" depends on the understanding of the person skilled in the art when applying the relevant measurement processes. By way of example, a beam profile is under- stood to be an intensity distribution along a cross section through the particle beam perpendicular to its direction of propagation.

Further, the test structure can have (slightly) angled flanks, i.e., can deviate from an exact rectangular geometry in this respect. By way of example, such a deviation may be caused by an opening angle of the electron beam which is used to generate the test structure by means of an electron beam-induced process. By way of example, the opening angle is between 0.1° and 5°, preferably less than or equal to 2°, preferably less than or equal to 1°, further preferably less than or equal to 0.5°. This opening angle can be reflected in a corresponding flank angle. In this case, the flank angle can also be greater than the opening angle of the electron beam.

In further embodiments, the at least one test structure has a curved profile in a plan view. By way of example, curved means that the first direction at a first po- sition is a different direction to the first direction at a second position that differs from the first position. A radius of curvature of the curved profile is preferably greater than a width of an elevation or depression, in particular more than twice the radius of curvature and up to 20 times the radius of curvature. The elevations and depressions can have for example an arc-shaped embodiment, in particular a circular arc-shaped embodiment. The parallelism criterion remains observed at each point of the test structure in this case.

In further embodiments, the test structure has a regular arrangement of cubes or cuboids. In the plan view, this test structure would resemble a chequerboard, for example. This test structure corresponds to a superposition of two line-like test structures with an orthogonal alignment at the same position. This embodiment is particularly space saving.

According to a further embodiment of the method, a length of a front section of the measuring tip, the diameter of which is less than a width of the depression of the test structure, is determined in step c).

According to a further embodiment of the method, a structure size of a sample to be measured or manipulated with the aid of the measuring tip is ascertained prior to step a) and the at least one test structure is generated in step a) on the basis of the ascertained structure size.

This embodiment is advantageous if the emphasis lies in ascertaining whether a specified limit structure size is still able to be measured by the respective meas- uring tip. By way of example, if the measuring tip capturing a structure with an aspect ratio of 1:1 and a height of 25 nm without errors suffices in one application, then it may be sufficient to generate a test structure with exactly these geometric properties and scan the latter with the measuring tip.

In particular, the measuring tip is set up to measure and/or manipulate a micro- lithographic photomask.

According to a further embodiment of the method, the test structure generated in step a) is removed again after the geometry of the measuring tip has been ascer- tained.

Preferably, the test structure is removed in this case by a precursor gas suitable to this end by means of a particle beam-induced etching process. By way of exam- ple, the precursor gas may comprise: xenon difluoride (XeF ₂ ), xenon dichloride (XeCl ₂ ), xenon tetrachloride (XeCl ₄ ), steam (H ₂ O), heavy water (D ₂ O), oxygen (O ₂ ), ozone (O ₃ ), ammonia (NH ₃ ), nitrosyl chloride (NOCl) and/or one of the fol- lowing halide compounds: XNO, XONO ₂ , X ₂ O, XO ₂ , X ₂ O ₂ , X ₂ O ₄ , X ₂ O ₆ , where X is a halide. Further etching gases for etching one or more of the deposited test structures are specified in the applicant’s US patent application having the num- ber 13/0 103 281.

Such an etching process can advantageously be carried out in situ, i.e., without a vacuum having to be broken and/or the sample having to be moved into another device.

According to a second aspect, a method for analysing and/or processing a sample, in particular a microlithographic photomask, is proposed. The sample is analysed and/or processed in a first step a) with the aid of a measuring tip of a scanning probe microscope. A geometry of the measuring tip is determined according to the method according to the first aspect in a step b), which can be carried out before and/or after step a).

According to one embodiment of the method, steps a) and b) are implemented in the same vacuum chamber and/or without the vacuum being broken. According to a further embodiment of the method, the at least one test structure on the sample and/or a substrate is generated within a vacuum chamber of the scanning probe microscope in step b).

According to a further embodiment of the method, the sample is processed by means of a particle beam-induced process with the aid of a focussed particle beam, in particular with an electron beam, and at least one precursor gas or etch- ing gas.

The sample being processed by the particle beam-induced process is understood to mean that, in particular, a deposition process and/or an etching process is car- ried out at a processing position of the sample, for example a defect of a lithogra- phy mask. For these processes, it is possible to use the same process gases as for the generation of the test structure described above.

In particular, processing of the sample and generating the test structure are im- plemented in the same vacuum chamber by means of the same apparatus, i.e., for example, the same electron beam unit and/or gas supply.

According to a third aspect, a test structure for determining a geometry of a measuring tip of a scanning probe microscope is proposed. The test structure has elevations alternating with depressions in a first direction, wherein the eleva- tions and depressions are aligned parallel to one another in a second direction perpendicular to the first direction.

In particular, the test structure has a target structure size and at least one de- pression.

The embodiments and features of the test structure described within the scope of the first aspect apply accordingly to the proposed test structure, and vice versa.

In particular, the test structure has elevations and depressions, the widths of which range from 1 to 50 nm and the heights of which range from 1 to 100 nm. One could also say that the test structure has a structure at spatial frequencies ranging from lO/μm to 100O/μm, which can also be referred to as a resolution of the test structure. The test structure can be embodied in such a way that the res- olution varies two-dimensionally in space, i.e., becomes higher/lower along the first direction, for example. The function which describes the variation is for ex- ample a step function with the piecewise constant resolution or else a continuous function. By way of example, the resolution can increase linearly or else increase logarithmically or exponentially and/or the function is a combination of different components.

The test structure can be embodied in such a way that an aspect ratio of the ele- vations and depressions varies along the first direction. By way of example, in this case a respective elevation/depression pair can have a certain first aspect ra- tio and the subsequent elevation/depression pair can have a different aspect ra- tio.

The test structure can further be embodied in such a way that the structure size of the elevations and depressions varies with a constant aspect ratio along the first direction. In this case, a respective elevation/depression pair can have a spe- cific first structure size with a certain aspect ratio and the following elevation/de- pression pair can have a different structure size with the same specific aspect ra- tio.

Such a variation of the aspect ratio or the structure size can be implemented ac- cording to a step function or else continuously according to a continuous function. Preferably, sections with the same constant geometry of the test structure in each case are greater than a lower limit along the first direction. By way of exam- ple, the lower limit can be determined in such a way that a respective section comprises at least ten elevation/depression pairs.

The test structure is preferably produced by means of a particle beam-induced process, in particular by an electron beam in conjunction with a suitable precur- sor gas. In this case, it is possible to make use of etching processes, which ablate material, and/or deposition processes, which apply material. In embodiments, the test structure can furthermore have a plurality of rounded, for example circular, column-like elevations and/or depressions, which are prefer- ably arranged in regular fashion and have a certain aspect ratio.

According to a fourth aspect, the use of a test structure according to the third as- pect is proposed for determining a geometry of a measuring tip of a scanning probe microscope.

According to a fifth aspect, a test device for determining a geometry of a measur- ing tip for a scanning probe microscope is proposed. The test device comprises a generating unit for generating at least one test structure, which has elevations alternating with depressions in a first direction, wherein the elevations and de- pressions are aligned parallel to one another in a second direction perpendicular to the first direction. Furthermore, the test device has a scanning unit for scan- ning the test structure with the measuring tip and for ascertaining a profile of the test structure. A determining unit of the test device is set up to determine the geometry of the measuring tip on the basis of the ascertained profile.

In particular, the test structure is generated with a target structure size and has at least one depression. In particular, the determining unit is configured to deter- mine the geometry of the measuring tip on the basis of the ascertained profile and the target structure size and/or an ascertained actual structure size.

The embodiments and features described for the proposed method according to the first aspect are correspondingly applicable to the proposed test device.

The test device is advantageously set up to carry out the method according to the first aspect.

In particular, the generating unit is set up to generate a test structure according to the third aspect.

The determining unit can be implemented in terms of hardware and/or in terms of software. In the case of an implementation in terms of hardware, the determin- ing unit can be embodied as a computer or as a microprocessor, for example. In the case of an implementation in terms of software, the determining unit can be embodied as a computer program product, as a function, as a routine, as an algo- rithm, as a neural network, as part of a program code or as an executable object.

According to a sixth aspect, a device for analysing and/or processing a sample, in particular a microlithographic photomask, is proposed. The device comprises an analysing and/or processing unit for analysing and/or processing the sample with the aid of a measuring tip of a scanning probe microscope. Furthermore, the de- vice comprises a test device according to the fifth aspect, wherein the test device is set up to determine a geometry of the measuring tip before and/or after the analysis and/or processing of the sample.

The embodiments and features described for the proposed method according to the second aspect are correspondingly applicable to the proposed device.

The device is advantageously set up to carry out the method according to the sec- ond aspect.

According to one embodiment of the device, the latter further comprises a pro- cessing unit for processing the sample by means of a particle beam-induced pro- cess with the aid of a focussed particle beam, in particular with an electron beam, and at least one precursor gas or etching gas.

If the test structure is likewise generated by means of a particle beam-induced process, the particle beam for processing the sample and for generating the test structure is provided by the same particle beam providing unit and the utihzed process gas is preferably supplied from the same process gas providing unit. Fur- ther, the same process parameters, such as a gas composition, in particular par- tial gas pressures, can be used for generating the test structure and processing the sample.

“A(n); one” in the present case should not necessarily be understood as restrictive to exactly one element. Rather, a plurality of elements, such as, for example, two, three or more, can also be provided. Any other numeral used here, too, should not be understood to the effect that there is a restriction to exactly the stated number of elements. Rather, numerical deviations upwards and downwards are possible, unless indicated to the contrary.

Further possible implementations of the invention also comprise not explicitly mentioned combinations of features or embodiments that are described above or below with respect to the exemplary embodiments. In this case, a person skilled in the art will also add individual aspects as improvements or supplementations to the respective basic form of the invention.

Further advantageous configurations and aspects of the invention are the subject matter of the dependent claims and also of the exemplary embodiments of the in- vention described below. In the text that follows, the invention is explained in more detail on the basis of preferred embodiments with reference to the accompa- nying figures.

Fig. 1a shows a schematic view of a measuring probe for a scanning probe micro- scope;

Fig.1b shows three different, measuring tips with different geometries;

Fig. 2 shows a schematic view of a first embodiment of a test structure;

Fig. 3 schematically shows scanning of a test structure with a measuring tip and a resultant profile;

Fig. 4 shows a schematic electron microscope image of a sample with a plurality of test structures;

Fig. 5 shows cross sections through two test structures with different aspect ra- tios;

Fig. 6 shows a sequence with three test structures with decreasing aspect ratio and the corresponding profiles; Fig. 7 shows an electron microscope image of a test structure and the correspond- ing profile;

Fig. 8 shows a diagram in which the penetration depth of a measuring tip is illus- trated as a function of a structure size of the test structure in exemplary fashion;

Fig. 9 shows an arrangement of a plurality of different test structures;

Fig. 10 shows three further different, embodiments of test structures;

Fig. 11 schematically shows a sample to be analysed;

Fig. 12 shows an exemplary block diagram of an exemplary embodiment of a de- vice for analysing and processing a sample using a test device;

Fig. 13 shows a schematic block diagram of an exemplary embodiment of a method for determining a geometry of a measuring tip;

Fig. 14 shows a schematic view of a measuring tip which partly dips into a de- pression of a test structure; and

Fig. 15 shows a schematic view of a cross section of a test structure which was generated by means of an electron beam-induced deposition process.

Identical elements or elements having an identical function have been provided with the same reference signs in the figures, unless indicated to the contrary. It should also be noted that the illustrations in the figures are not necessarily true to scale. Figure 1a shows a schematic view of a measuring probe 110 for a scanning probe microscope 514 (see Figure 12). The measuring probe 110 has a holding beam 120, referred to as a cantilever in the art, which is arranged on a fastening sec- tion 130. Arranged at the free end of the cantilever 120 is the measuring tip 100. In this example, the measuring tip has two sections 101, 102. The front section 101 has a particularly fine embodiment. By way of example, the front section 101 has a length of 200 nm with a diameter of 20 nm. Alternatively, the measuring tip 100 has no such subdivision into various sections with different geometries.

The cantilever 110 and the measuring tip 100 may be configured in one piece. By way of example, the cantilever 120 and the measuring tip 100 may be manufac- tured from a metal, such as, for instance, tungsten, cobalt, iridium, a metal alloy, or a semiconductor, such as, for instance, silicon, silicon nitride, silicon carbide or silicon oxide. Preferably, the measuring tip 100 is grown on the cantilever 120 by means of a particle beam-induced process. Moreover, the measuring tip 100 can be manufactured from different materials, for example have a base made of a first material and a tip made of a second material. It is also possible to manufac- ture the cantilever 110 and the measuring tip 100 as two separate components and to subsequently connect these to one another. This can be effectuated by ad- hesive bonding, for example.

Figure 1b shows three different measuring tips 100a, 100b, 100c with different geometries in exemplary fashion. The left column of Figure 1b illustrates a side view of the respective measuring tip and the right column of Figure 1b illustrates the cross section through the respective measuring tip at different positions.

The first measuring tip 100a has, e.g., a conical geometry. The latter tapers coni- cally to the tip, with the cross section, as shown by the exemplary cross sections C1, C2, C3, being circular in each case.

The second measuring tip 100b has, e.g., a pyramid-like geometry with a triangu- lar base C1, C2, C3. The respective corner of the triangular base forms a flank F1, F2, F3 of the measuring tip 100b. The flank F1 extends substantially perpen- dicular in relation to the cross sections C1, C2, C3; the two flanks F2, F3 have a tilt. Hence, this measuring tip 100b is suitable for capturing perpendicular edges of a structure if the measuring tip 100b is made to approach the structure with the flank F1.

The third measuring tip 100c has a cylindrical profile, with the cross section C1 being substantially constant along the length of the measuring tip 100c. Figure 2 shows a schematic perspective view of a first embodiment of a test struc- ture 200 with a target structure size B1, B2, H1, H2 (see Figure 4 or 5). In this example, the test structure 200 is arranged on a substrate 230, which might be a semiconductor substrate for example. The test structure 200 has a line-like geom- etry in a plan view, with a comb-like cross section. In this case, elevations 210 are arranged alternately with depressions 220 in a first direction I. The elevations 210 and depressions 220 extend parallel to one another in a second direction II, which is perpendicular to the first direction I.

In this example, the test structure 200 was generated by a deposition of material at the positions of the elevations 210. One could say that the elevations 210 were grown on. In this case, growth was implemented along the third direction III, which is perpendicular to the first direction I and the second direction II. By way of example, the third direction III is parallel to a surface normal (not shown) of the substrate 230 in this case. It should be observed that Figure 2 represents the test structure 200 with the target structure size B1, B2, H1, H2. The generated test structure 200 can have deviations therefrom. Figure 15 shows an example of a test structure 200 with such deviations, which may be caused by a limited pro- cess accuracy of the generation process.

In this example, the elevations 210 and depressions 220 have a uniform design, i.e., the width of an elevation 210 and the width of a depression 220 are the same in each case. By way of example, the width is 10 nm and the height of the eleva- tions 210 above the substrate is 30 nm. Hence, the elevations 210 and the depres- sions 220 have an aspect ratio of 3:1.

Figure 3 schematically shows scanning of a test structure 200 with a measuring tip 100 and a resultant profile 300 therefrom. By way of example, this relates to the measuring tip 100a of Figure 1b with a conical geometry and to the test structure 200 of Figure 2. The left side of Figure 3 schematically illustrates how the measuring probe 110, which is held by a probe holder 140 (see Figure 12), is raster scanned over the test structure 200 in order to scan the latter. This is an atomic force microscope in this example. The measuring probe 110 is moved back-and-forth parallel to the first direction I, with a control loop (not illustrated) keeping the interaction of the measuring tip 100 with the surface of the test structure 200 constant by virtue of adjusting the probe holder 140 with the measuring probe 110 in the third direction III. Sche- matically and for easier understanding, the measuring tip 100 is shown in con- tact with the test structure 200 in Figure 3; however, this need not necessarily be the case in other modes of operation of the atomic force microscope. At the illus- trated time, the measuring tip 100 is in a depression 220 of the test structure 200, with a front flank of the measuring tip 100 just coming into contact with an edge between the depression 220 and the subsequent elevation 210. If the meas- uring probe 110 is now advanced further, the probe holder 140 needs to be moved upward in the third direction III. In this case, an angle of the trajectory of the probe holder 140 in relation to the horizontal direction corresponds to the angle of the front flank of the measuring tip 100 in relation to the horizontal direction.

Therefore, the profile 300 illustrated in the right-hand diagram arises as the re- sult of scanning the test structure 200. Here, the dashed lines represent the ac- tual geometry (target structure size) of the test structure 200. From this example, it is evident that the profile 300 ascertained by scanning does not correspond to the structure actually present. Rather, the profile 300 is a convolution of the ge- ometry of the test structure 200 and the geometry of the measuring tip 100. If one or another geometry is known in this case, the respective other geometry can be determined by deconvolution of the profile 300. Thus, it is possible to deter- mine the geometry of the measuring tip 100 if the geometry of the test structure 200 is known.

In particular, the geometry of the measuring tip 100 is determined by virtue of the ascertained profile 300 being analysed or assessed on the basis of the target structure size B1, B2, H1, H2 (see Figure 5) and/or the actual structure size B1*, B2*, H1* (see Figure 4, 7 or 15). That is to say, the target structure size B1, B2, H1, H2 and/or the actual structure size B1*, B2*, H1* is included when deter- mining the geometry of the measuring tip 100.

As a result of an advantageous design of the test structure 200 this allows vari- ous geometry parameters of the measuring tip 100 to be determined, for example a length, a diameter as a function of the length, a work angle, a flank of the measuring tip in different directions, and the like. In the present case, determin- ing the geometry of the measuring tip is understood to mean that at least one of the geometry parameters is determined.

Figure 4 shows a schematic electron microscope image 400 of a sample 10 (see Figure 11 or 12) with a plurality of different test structures 200a-d. In this exam- ple, four different test structures 200a-d are illustrated. All four test structures 200a-d have a line-shaped structure with a comb-like cross section, as explained in detail on the basis of Figure 2. By way of example, the hatched bars represent elevations 210 (see Figure 2) or, alternatively, they can also relate to depressions 220 (see Figure 2). Without loss of generality, the test structures 200c, 200d in this example each have three elevations/depressions and the test structures 200a, 200b each have four elevations/depressions. The test structures 200a, 200c have the same alignment and the test structures 200b, 200d have an alignment orthogonal thereto. Moreover, the test structures 200a, 200b have a smaller structure size than the respective corresponding test structures 200c, 200d. As an example for the structure size, the width B1 of an elevation 210 and the width B2 of a depression 220 are illustrated on the test structure 200a. One could say that the test structures 200a and 200c have a different spatial resolution or have in- formation at different spatial frequencies. The same is true for the test structures 200b and 200d. Advantageously, the actual structure size B1*, B2* and/or the resolution of a respective test structure 200a-d can be ascertained from the elec- tron microscope image 400.

The different resolution of the test structures 200a, 200c or 200b, 200d allows dif- ferent geometry parameters of a measuring tip 100 (see Figure 1a, 1b or 12) to be ascertained, or certain geometry parameters can be ascertained with different ac- curacies. Moreover, the geometry of the measuring tip 100 can be ascertained more accurately in relation to various flanks F1-F3 (see Figure 1b) as a result of the orthogonal alignment of in each case two of the test structures 200a-d. The respective test structure 200a-d is preferably scanned perpendicular to the re- spective line direction. By way of example, a front and a rear flank are decisive in the case of the test structures 200a, 200c, and so the geometry of said flanks can be determined from the profile 300 (see Figure 3) ascertained here, and the left and right flank are decisive in the case of the test structures 200b, 200d, and so the geometry of said flanks can be determined from the profile 300 ascertained there.

Figure 5 shows cross sections through two test structures 200a, 200b with differ- ent target structure sizes B1, B2, H1, H2. In this example, the different target structure sizes B1, B2, H1, H2 lead to different aspect ratios AV1, AV2. The two test structures 200a, 200b have a comb-like cross section. The aspect ratio AV1, AV2 is the ratio of the height H1, H2 of the respective depression 220 to the width B1, B2 of the respective depression 220. In this example, the aspect ratio AV1 is greater than the aspect ratio AV2. The respective aspect ratio can be spec- ified accordingly for the further elevations 210 or further depressions 220 of the respective test structure 200a, 200b. In this example, the test structures 200a, 200b have a uniform embodiment, i.e., the aspect ratio AV1, AV2 is the same for each elevation 210 or depression 220.

Figure 6 shows a sequence with three test structures 200a-c with different target structure sizes B1, B2, H1, H2 (see Figure 4 or 5), in particular with a decreasing aspect ratio AV1, AV2 (see Figure 5) and corresponding profiles 300a-300c. In this example, the respective test structure 200a-c has a line-shaped structure with a comb-like cross section. In this case, the regions illustrated with hatching correspond to elevations 210 (see Figures 2, 3, 5). In this case, these have a height H1, H2 (see Figure 5) of 30 nm. The aspect ratio of the left test structure 200a is, e.g., 3:1, i.e., the elevations 210/depressions 220 have a width B1, B2 (see Figure 5) of 10 nm. The aspect ratio of the central test structure 200b is, e.g., 6:1, i.e., the elevations 210/depressions 220 have a width B1, B2 of 5 nm. The aspect ratio of the right test structure 200c is, e.g., 10:1, i.e., the elevations 210/depres- sions 220 have a width B1, B2 of 3 nm.

The three illustrated test structures 200a-c are scanned sequentially with in- creasing aspect ratio using a measuring tip 100 (see Figure 1a, 1b, 3 or 12) in or- der to obtain a corresponding profile 300a-c. In this case, the scanning is imple- mented along the first direction I of the respective test structure 200a-c. By way of example, the profiles 300a-c illustrated in the diagrams are obtained in this case. The horizontal axis I corresponds to a position of the respective test structure 200a-c along the first direction I, the vertical axis III corresponds, e.g., to a height value, at which the interaction of the measuring tip with the surface of the respective test structure 200a-c is constant. The left test structure 200a is captured in its entirety by the measuring tip 100, in particular the measuring tip 100 reaches the base of the depressions 210. A slight work angle of the measuring tip 100 may be identifiable in the profile 300a at the edges between an elevation 210 and a depression 220. In the central test structure 200b, the measuring tip 100 only just still reaches the base of a depression 220 between two elevations 210, as is identifiable by a short flat section in the profile 300b. By contrast, in the right test structure 200c, the measuring tip 100 only penetrates a small dis- tance into the depressions 220 and the base of the depressions 220 is no longer reached.

It is immediately identifiable from these three measurements that this measur- ing tip 100 is not suitable for the analysis of depressions with a diameter of 5 nm and an aspect ratio of 10: 1. The structure size of the central test structure 200b could be specified as limit of the suitability of the measuring tip 100. A geometry of the respective flank F1-F3 (see Figure 1b) of the measuring tip 100 can be de- duced from the rising flanks of the profiles 300a-c. In particular, what can be de- rived from the profile 300c is that the measuring tip 100 has a diameter of 5 nm at the maximum penetration depth since this is the width of the depressions 220. The penetration depth of the measuring tip 100 can be presented as a function of the width of the depressions 220, as illustrated in exemplary fashion in Figure 8.

Figure 7 shows an electron microscope image of a test structure 200 and the cor- responding profile 300, which was captured by scanning the test structure 200 with a measuring tip 100 (see Figures 1a, 1b, 3, 12). In this example, the test structure 200 has a line-shaped structure with a comb-like cross section. The bright lines identifiable in the electron microscope image correspond to the eleva- tions 210, the dark regions therebetween correspond to the depressions 220. By way of example, the test structure 200 has an aspect ratio of 3A, with a height of the elevations being 60 nm, for example. By way of example, the test structure 200 was constructed by means of an electron beam using tetraethoxysilane as deposition gas. The actual structure size B1*, B2* of the test structure can be ad- vantageously ascertained from the illustrated electron microscope image. By way of example, the elevations 210 have an actual width B1* of 20 nm and the depres- sions 220 have an actual width B2* of 60 nm. The test structure 200 was scanned using an atomic force microscope and the profile 300 was captured thus. The pro- file 300 shows rounded edges of the elevations 210 and downwardly tapering de- pressions 220. Such a profile 300 is referred to as substantially rectangular in the present case. As already explained, the corners of the test structure 200 are fre- quently slightly rounded on account of a Gaussian beam profile, for example (see also Figure 15). This becomes more noticeable with narrower elevations and de- pressions. The profile 300 shows the measuring tip 100 reaching the substrate between the elevations 210. What is derivable herefrom is that the measuring tip 100 has a diameter of less than 60 nm in its front section 101 (see Figure 1a).

Figure 8 shows a diagram in which the penetration depth ΔH of a measuring tip 100 (see Figure, la, 1b, 3 or 12) is represented by way of example as a function of a target structure size B1 of the test structure 200 (see Figures 2-7, 9, 10). By way of example, the horizontal axis B1 refers to a width B1 (see Figure 5) of the depression 220 (see Figures 2, 3, 5) of the test structure 200. By way of example, twelve test structures 200 (each point in the diagram corresponds to a test struc- ture) with in each case different, widths B1 of the depressions 220 but with a con- stant height H1 (see Figure 5) were generated and scanned using the measuring tip 100 in this case. A maximum penetration depth AH of the measuring tip was ascertained from the profiles 300 obtained in the process (see Figure 3, 6) and plotted in the diagram illustrated here. By way of example, the height H of the depressions 220 is 70 nm. The aspect ratio could also have been plotted on the horizontal axis instead of the width B1.

It is possible to identify that the penetration depth AH is substantially 0 up to a width B1 of 20 nm. From this, it is possible to deduce that the measuring tip 100 has a diameter of approximately 20 nm at its front end. The penetration depth AH then increases continuously, wherein the diameter of the measuring tip 100 at a distance from the front end of the measuring tip 100 that corresponds to the penetration depth AH can be read from a respective value. This is illustrated in exemplary fashion for the depression 220 with a width B1 of 40 nm. Conse- quently, the measuring tip 100 has a diameter of 40 nm in the case of a penetra- tion depth ΔH of 20 nm. From a width B1 of 70 nm, the measuring tip 100 reaches the base of the depression 220 in the case of a penetration depth ΔH of 70 nm.

Figure 9 shows an arrangement of a plurality of different test structures 200a-f. The test structures 200a-f are line-shaped structures with a comb-like cross sec- tion. The test structures 200a, 200c and 200e have the same alignment with dif- ferent structure sizes or different aspect ratios. The test structures 200b, 200d and 200f are aligned orthogonal thereto and have corresponding different struc- ture sizes or aspect ratios. Figure 9 is an example of an arrangement of a plural- ity of test structures 200a-f with different alignments and a variation in the as- pect ratio.

Figure 10 shows three further examples of embodiments of test structures 200a- c. All three test structures 200a-c comprise elevations 210 alternating with de- pressions 220 in a first direction I, wherein the elevations 210 and depressions 220 are aligned parallel to one another in a second direction II perpendicular to the first direction I.

In the case of the test structure 200a, the elevations 210 and depressions 220 have a circular arc shape, wherein a respective circular arc spans an angle of 90°. Moreover, in the radial direction, the test structure 200a has a variation in the structure size, in particular the width B1, B2, which is plotted in exemplary fash- ion for two of the circular arcs. The elevations 210 and depressions 220 embodied as circular arcs have a common centre. A tangent T is respectively plotted at three edges between an elevation 210 and a depression 220. In this case, the re- spective tangent T relates to a point of the respective circular arc which corre- sponds to a point of intersection with a line I extending in the radial direction. The elevations 210 and depressions 220 are aligned parallel to one another in re- lation to this radial direction. Consequently, the radial direction corresponds to the first direction I and the tangential direction corresponds to the second direc- tion II.

In addition to circular arc-shaped structures, the elevations 210 and depressions 220 can have further curvy or curved profiles. The test structure 200b has a chequerboard-like arrangement of elevations 210 and depressions 220. This embodiment unifies two mutually orthogonally ar- ranged test structures 200 in a small area. In this example, the structure size is uniform but a variation in the structure size is also possible, for example a line- by-line increase or reduction in the structure size or the aspect ratio.

The test structure 200c has a line-shaped structure with a comb-like cross sec- tion, wherein a width B1 of the depressions 220 is gradually reduced along the first direction I. This provides the option of ascertaining a limit size, which is still able to be captured without problems by the respective measuring tip 100, using a single scanning procedure. This embodiment can therefore contribute to a faster determination of the geometry of the measuring tip 100.

Figure 11 schematically shows a sample 10 to be analysed, for example a section of a structured lithography mask. In particular, the lithography mask 10 has a structure formed by a coating 14 on the substrate 12 of the sample 10. The struc- ture size B of these structures can be different at various positions of the lithog- raphy mask 10. By way of example, the width B of a region is plotted as structure size in Figure 11. By way of example, the structure size B lies in a region of 20- 200 nm. Occasionally, defects D arise during the production of lithography masks, for example because etching processes do not run exactly as intended. In Figure 11, such a defect D is illustrated with hatching. This is excess material since the coating 14 was not removed from this region even though the two coat- ing regions 14 next to one another are envisaged as separate in the template for the lithography mask 10. One could also say that the defect D forms a web. In this case, a size of the defect D corresponds to the structure size B. Other defects which are smaller than structure size B, for example lying in the region of 5- 20 nm, are also known. Such defects D, and also other types of defects, can be rec- tified in devices that are suitable to this end. In this example it is necessary to re- move the web in a targeted manner, for example by particle beam -induced etch- ing. In so doing, it is advantageous if the defect site is monitored using imaging processes before, during and/or after the etching process. In particular, a scan- ning probe microscope, preferably an atomic force microscope, is suitable for ob- taining a three-dimensional image. As described above on the basis of Figure 3, the geometry of the utilized measur- ing tip 100 (see Figures la, 1b, 3, and 12) should be known as accurately as possi- ble for an exact interpretation of the measurement data obtained in the process, the profile 300 (see Figures 3 and 6). To this end, one of the above-described test structures 200 (see Figure 2-7, 9, 10 or 12) can be advantageously used. Depend- ing on the type of defect D and how the latter is rectified, it may be sufficient to ensure that the measuring tip 100 does not exceed a certain limit, in respect of its width. By way of example, in the example of Figure 11, it may be sufficient to de- termine that, after the defect D has been removed, the substrate 12 is reached along a narrow line and the two substrate regions 12 above and below the illus- trated defect D are connected. Therefore, it may be sufficient to generate the test structure 200 in such a way that the latter has, e.g., a structure size like the de- fect D, i.e., in particular a width B as illustrated here. The height of the coating 14 is defined and constant for the lithography mask 10, and so an aspect ratio AV1, AV2 (see Figure 5) is set at the same time as the width B. One could there- fore say that the test structure 200 is generated on the basis of or depending on the structure size B of the sample 10. Hence, determining the geometry of the utilized measuring tip 100 can be restricted to one or a few test structures 200, which may be accompanied by saving of time. The analysis of the defect D or of the defect site after the defect D has been removed is hence likewise accelerated and possible with a greater reliability.

Fig. 12 shows an exemplary block diagram of an exemplary embodiment of a de- vice 500 for analysing and processing a sample 10 using a test device 510. The device 500 is largely arranged in a vacuum housing 501, which is kept at a cer- tain gas pressure by a vacuum pump 502.

The device 500 is set up in particular for analysing and processing lithography masks 10 (see Figure 11). By way of example, this is a verification and/or repair tool for lithography masks, in particular for lithography masks for EUV ("ex- treme ultraviolet") or DUV ("deep ultraviolet") lithography. A sample 10 to be an- alysed or processed is mounted on a sample stage 11 in this case. In particular, the sample stage 11 is set up to set the position of the sample 10 in three spatial directions and in three axes of rotation with an accuracy of a few nanometres.

The device 500 comprises an electron column 540. The latter comprises an electron source 541 for providing an electron beam 542 and an electron micro- scope 543, which captures the electrons backscattered from the sample 10 in this arrangement. A further detector for secondary electrons may also be provided (not illustrated). The electron column 540 can carry out electron beam-induced processing (EBIP) processes in conjunction with supplied process gases, which are supplied into the region of a focal point of the electron beam 542 in the sam- ple 10 from the outside via a valve 532 and a gas line 534 by way of a gas provid- ing unit 530. In particular, said processes comprise a deposition and/or an etch- ing of material. Together with the gas providing unit 530, the electron column 540 can be referred to as a generating unit 512, for example. It is observed that the electron microscope 543 is not mandatory for EBIP, but it is advantageous for monitoring the processes.

The device 500 furthermore comprises an analysing and processing unit 520, which is embodied as an atomic force microscope in this case. The atomic force microscope 520 comprises a scanning unit 514, which has a probe holder 140 that is displaceable in at least three spatial directions by means of a plurality of piezo actuators (not shown). Moreover, the probe holder 140 can be mounted so as to be rotatable about one or more axes in order to compensate a tilt or bring this about in a targeted manner and/or in order to adjust a relative goniometric alignment in relation to the sample 10. In particular, the probe holder 140 is set up to re- ceive a measuring probe 110 (see Figure 1) having a cantilever 120 and a measur- ing tip 100. The probe holder 140 is set up to raster scan a surface of the sample 10 using the measuring tip 100. To this end, the scanning unit 514 preferably comprises a closed control loop (not shown), which for example keeps an interac- tion of the measuring tip 100 with the sample 10 constant. In particular, this cap- tures a height profile of the sample surface. To obtain the height profile from the raw measurement data, it is necessary, for example, to make an assumption about the geometry of the measuring tip 100.

In order to be able to determine the geometry of the measuring tip 100 in situ, it is possible to generate a test structure 200 with a specified geometry, in particu- lar with a predetermined aspect ratio, by means of the generating unit 512. In this case, the test structure 200 can be embodied as described on the basis of Fig- ure 2-7, 9 or 10. In this example, the test structure 200 is arranged on the sample 10 itself. Here, the test structure 200 can be arranged both on the substrate 12 (see Figure 11) or on the coating 14 (see Figure 11) of a lithography mask. The test structure 200 is preferably embodied in such a way that it can be removed without residue in a subsequent process step. As an alternative to the position of the sample 10, the test structure 200 can be generated on a position of the sam- ple stage or of a separate sample (not shown) especially provided to this end.

The device 500 moreover comprises a determining unit 516, which is embodied in this case as a constituent part of a control computer 503 arranged outside of the vacuum housing 501. In particular, the determining unit 516 is set up to deter- mine the geometry of the measuring tip 100 on the basis of a profile 300 (see Fig- ures 3 and 6) of the test structure 200 captured by the measuring tip 100. The control computer 503 is set up to control the electron column 540 and the atomic force microscope 520.

The atomic force microscope 520 can also be used as a micromanipulator. In this case, for example, the measuring tip 100 is moved to a position on the sample 10 to be manipulated and brought into contact with the sample there, for example.

In this way, it is possible, for example, for the measuring tip 100 to gather the dirt adhering to the sample surface, such as dust particles, and consequently re- move the latter.

The test device 510 comprises at least the generating unit 512, the scanning unit 514 and the determining unit 516.

Figure 13 shows a schematic block diagram of an exemplary embodiment of a method for determining a geometry of a measuring tip 100 (see Figure 1a, 1b, 3 or 12). At least one test structure 200 (see Figures 2-7, 9, 10), which has eleva- tions 210 (see Figure 2, 3, 5 or 10) alternating with depressions 220 (see Figure 2, 3, 5 or 10) in a first direction I (see Figure 2, 3 or 10) is generated in a first step SI, wherein the elevations 210 and depressions 220 are aligned parallel to one another in a second direction II (see Figures 2 and 10) perpendicular to the first direction I. By way of example, the test structure 200 has one of the structures il- lustrated in Figure 2-7, 9 or 10. The test structure 200 is preferably generated by the generating unit 512 (see Figure 12) of the test device 510 (see Figure 12) or of the device 500 (see Figure 12). The test structure 200 is scanned in a second step S2 by the measuring tip 100 for ascertaining a profile 300 (see Figure 3 or 6) of the test structure 200. In particular, this is implemented by means of the scan- ning unit 514 (see Figure 12) of the test device 510 or of the device 500. In a third step S3, the geometry of the measuring tip 100 is ascertained on the basis of the ascertained profile 300. By way of example, this is implemented by the determin- ing unit 516 (see Figure 12) of the test device 510 or of the device 500.

If the geometry of the measuring tip 100 is determined, the measuring tip 100 can be used to analyse and/or process a sample 10. In particular, a suitability of the measuring tip 100 for a certain analysis task or processing task is known in that case. Profiles 300 ascertained with the measuring tip 100 can be converted with great reliability into height profiles of the scanned surface on the basis of the known geometry of the measuring tip 100.

Advantageously, a change in the geometry of the measuring tip, which may occur on account of wear or use, for example, can be identified by repeated scanning of the test structure 200 with the measuring tip 100, and so it is possible to monitor the actual geometry of the measuring tip 100 and possible to avoid incorrect measurements.

Furthermore, the generated test structure 200 can be removed again in situ, i.e., without the vacuum being broken, for example by means of the particle beam-in- duced etching process.

Figure 14 shows a schematic view of a measuring tip 100 which partly dips into a depression 220 of a test structure 200. In this example, the test structure 200 has only a single depression 220, which is flanked by two elevations 210. The pene- tration depth ΔH of the measuring tip 100 is limited since the measuring tip 100 makes contact with the test structure 200 at two points P1, P2 of a respective edge which marks the transition from the respective elevation 210 to a respective flank of the depression 220. One could also say that the test structure 200 is cap- tured or scanned simultaneously at two different points in this situation. The il- lustrated situation, in which the measuring tip 100 simultaneously makes con- tact with two points P1, P2 of the test structure 200, is obtained when the test structure 200 is scanned by the measuring tip 100 if, in particular, a scan incre- ment of the AFM (resolution of the raster points) is small in relation to the diam- eter DI of the measuring tip 100. The scan increment denotes the pitch of two ad- jacent measurement points, which are scanned by the AFM. The scan increment of the AFM is "small" in relation to the diameter DI of the measuring tip 100 if, for example, the diameter DI of the measuring tip 100 is five times or ten times the scan increment, or even more than this. From the profile of the test structure 200 that is ascertained during such a measurement, it is possible to ascertain the diameter DI of the measuring tip 100 at a position which is spaced apart from the front point of the measuring tip 100 by the penetration depth ΔH, as already ex- plained on the basis of Figure 8.

Figure 15 shows a schematic view of a cross section of a test structure 200 which was generated by means of an electron beam-induced deposition process. A cross section of a generated test structure 200 as illustrated in Figure 15 can be cap- tured for example by virtue of an electron microscope image of a fracture edge of the test structure 200 and/or a TEM recording of a TEM lamella of the test struc- ture 200 being captured. On account of various influencing factors during the process for producing the test structure 200, which are in particular random in nature and/or due to the physical limits of accuracy (by way of example, the beam profile of the electron beam cannot be arbitrarily steep on account of Heisenberg's uncertainty principle), the test structure 200 may have slight variations and the elevations 210 and depressions 220 may deviate from an exact rectangular form. By way of example, these deviations relate to rounding at the transition from a flank to a plateau of an elevation 210 or depression 220, and a not exactly per- pendicular profile of the respective flank. The actual structure size B1*, H1* can be ascertained on the basis of a captured image representation (e.g., electron mi- croscope image) of the test structure 200. This may be different, for each elevation 210 or depression 220 on account of the statistical deviations when generating the test structure 200. In order to specify the actual structure size B1*, H1* us- ing a single value despite the variations present, it is possible for example to de- termine a confidence interval on the basis of an ascertained distribution of the re- spective actual structure size B1*, H1*. In particular, the confidence interval is determined on the basis of a statistical moment of the distribution. Thus, it is possible to specify, for example, that the test structure 200 has depressions 220 with an actual width B1* of 15±0.5 nm, where 15 nm is a mean value of all deter- mined actual widths B1* and 0.5 nm is a mean deviation of all determined actual widths B1* from the mean value. By virtue of suitably choosing the process pa- rameters for generating the test structure 200 and/or by virtue of carrying out suitable process monitoring during the production, it is possible to generate the test structure 200 precisely and with few variations. By characterizing the gener- ated test structure 200, preferably in situ, it is possible to analytically capture the deviations from an ideal shape and take these into account accordingly when ascertaining the geometry of the measuring tip 100.

Although the present invention has been described on the basis of exemplary em- bodiments, it is modifiable in diverse ways.

LIST OF REFERENCE SIGNS

10 Sample

11 Sample stage

12 Substrate 14 Coating

100 Measuring tip

100a-c Measuring tip

101 Section

102 Section

110 Measuring probe

120 Cantilever

130 Fastening section

140 Probe holder

200 Test structure

200a-fTest structure

210 Elevation

220 Depression

230 Substrate

300 Profile

300a-c Profile

400 Electron microscope image

500 Device

501 Vacuum housing

502 Vacuum pump

503 Control computer 510 Test device

512 Generating unit

514 Scanning unit

516 Determining unit

520 Analysing and processing unit

530 Gas providing unit

532 Valve

534 Line

540 Electron column 541 Electron source

542 Electron beam

543 Electron microscope ΔH Penetration depth AV1 Aspect ratio AV2 Aspect ratio B Structure size B1 Target width B1* Actual width B2 Target width B2* Actual width C1 Section C2 Section C3 Section D Defect structure

DI Diameter H1 Target height H1* Actual height H2 Target height

I First direction

II Second direction

III Third direction

P1 Contact point

P2 Contact point

S1 Method step

S2 Method step

S3 Method step T Tangent