Sample tilt measurement in dual-beam device

Abstract

A method for determining sample tilt in a dual-beam metrology device having a first charged-particle beam and a second charged-particle beam arranged at an angle α relative to each other. The method comprises projecting, from the first direction using the first charged-particle beam, a pattern onto the sample surface. The projected pattern is imaged on the sample surface from the second direction using the second charged-particle beam. By analyzing corresponding pattern geometries of the projected and imaged patterns, a tilt angle of the sample surface is determined. A further method for calibrating the angle α in in a dual-beam device is provided. The disclosed techniques enable real-time tilt correction of images based on the imaging and milling processing modules during 3D tomography processes.

Description

[0001] SAMPLE TILT MEASUREMENT IN DUAL- BEAM DEVICE

[0002] CROSS-REFERENCE TO RELATED APPLICATIONS

[0003] The present application claims the priority of German patent application DE 102024 123 804.6, the whole content of which is incorporated by reference herein for all purposes.

[0004] TECHNICAL FIELD

[0005] Various examples generally relate to operating a dual-beam device, for example configured for implementing a slice-and-image process of a sample. Various examples specifically relate to methods for determining a sample tilt in a dual-beam device comprising a charged-particle beam imaging module and a charged-particle sample processing module arranged in a fixed angle relative to each other.

[0006] BACKGROUND

[0007] Semiconductor devices are becoming increasingly complex, with smaller feature sizes and multiple layers of interconnects. To ensure the quality and reliability of these devices, it is essential to characterize their three-dimensional (3D) structure with high precision. A popular technique for 3D characterization is the slice-and-image process, also known as semiconductor tomography, which involves iteratively removing thin layers of material using a focused ion beam (FIB) and imaging the exposed sample surface using a scanning electron microscope (SEM).

[0008] In a dual-beam system, the FIB and SEM are arranged at a fixed known angle relative to each other. The sample, usually a wafer or a specific region of interest (ROI) on the wafer, is placed on a stage that can be tilted and rotated to align the sample surface to be imaged with the SEM beam for imaging. Ideally, the sample surface should be perpendicular to the SEM beam to ensure optimal image quality and accurate 3D reconstruction. DE 10 2023 120 736 B3 discloses a dual beam system comprising an optical detector for measuring a height and / or tilt of a sample.

[0009] However, due to various factors such as stage misalignment, sample mounting errors, and inherent wafer non-flatness (wafer bow), the sample surface in the region to be imaged often is not perpendicular to the SEM beam. This sample misalignment, characterized by a sample tilt angle relative to the perpendicular orientation with respect to the imaging beam, can introduce distortions and inaccuracies in the acquired images, leading to errors in the reconstructed 3D model.

[0010] Existing methods for determining and correcting sample tilt often involve time-consuming manual adjustments or rely on assumptions about the sample geometry. These methods have limited accuracy and repeatability, especially for high-precision applications such as defect analysis and metrology.

[0011] SUMMARY

[0012] Accordingly, there is a need for improved techniques for determining sample tilt in a dual-beam device, which alleviate or mitigate at least some of the above-identified restrictions and drawbacks.

[0013] This need is met by the features of the independent claims. The features of the dependent claims define further advantageous examples.

[0014] In the following, the solution according to the present disclosure will be described with regard to the claimed methods for determining sample tilt and for determining the angle between the dual charged-particle beams in a dual-beam device, as well as with regard to the claimed dual-beam devices, wherein features, advantages, or alternative embodiments can be assigned to the other claimed objects and vice versa. In other words, the claims related to the dual-beam devices can be improved with features described in the context of the methods, and the methods can be improved with features described in the context of the dual-beam devices.

[0015] Various examples of the present disclosure are described with regard to a dual-beam metrology device or system, in particular comprising a Focused Ion Beam (FIB) module and a Scanning Electron Microscope (SEM) module, wherein it is to be understood that the disclosed techniques can be applied to other dual-beam metrology devices or systems comprising at least two beams suitable for processing and imaging a sample surface. For example, the disclosed techniques can be applied to systems combining different types of ion beams or other particle beams in conjunction with various marking and / or imaging modalities.

[0016] A method for determining tilt of a sample, in particular of a sample surface of a sample, in a dualbeam metrology system is provided. Tilt of a sample may refer to an orientation of the sample surface to be imaged. Tilt may be relative to a machine coordinate system or the charged-particle beam of the imaging module.

[0017] The method may be a computer-implemented method. The method may be performed by a dualbeam device or dual-beam metrology system, as described in the present disclosure. At least part of the method may be performed by a computing device or processor comprised in the dual-beam device.

[0018] A corresponding dual-beam device, or dual-beam metrology system, is provided, which is configured to perform any one or any combination of methods as described in the present disclosure.

[0019] The dual-beam metrology device comprises a first charged-particle beam arranged in a first direction towards the sample, and a second charged-particle beam arranged in a second direction towards the sample, wherein the first and second directions define an angle a relative to each other. In a step, a pattern is projected onto the sample surface - from the first direction using the first charged-particle beam. In another step, the projected pattern on the sample surface is imaged from the second direction using the second charged-particle beam. In another step, at least one tilt angle of the sample surface is determined based on the pattern and the imaged pattern.

[0020] Projecting the pattern onto the sample surface may comprise structuring the pattern on the sample surface, e.g., by modifying the sample surface. The projected pattern may define a known spatial relationship, such as a grid, array, or set of distances between pattern elements. In the imaged pattern, these spatial relationships appear distorted or modified due to the tilt of the sample surface. By analyzing the differences between corresponding spatial relationships in the projected and imaged patterns, particularly between corresponding pattern elements in the projected respectively imaged patterns, the tilt angle can be determined. In various examples, the tilt angle of the sample surface can be determined by comparing a geometric characteristic of the projected pattern with the imaged geometric characteristic of the imaged pattern. The projected pattern may define a reference geometry or spatial relationship, such as a grid, array, or set of distances between pattern elements. When this pattern is imaged on a tilted sample surface, the geometry of the pattern elements in the captured image appears altered or distorted compared to the reference.

[0021] By analyzing the differences in the geometry or spatial relationships between the projected and imaged patterns, the tilt angle can be inferred. In general, this analysis may comprise quantifying and interpreting differences in corresponding projected and imaged pattern geometry, such as changes in distances, angles, or ratios between pattern elements.

[0022] In various examples, the projected pattern may comprise one or more measurable geometric characteristics, such as distances, angles, or ratios between pattern elements, which have known values when the pattern is projected towards as sample surface, or onto a non-tilted, perpendicular, surface. When the sample surface is tilted, these geometric characteristics deviate in the imaged pattern. By analyzing these deviations and relating them to the system geometry, the tilt angle can be estimated.

[0023] For example, the projected pattern may define a spatial encoding or mapping of positions on a non-tilted sample surface, and this encoding or mapping may be modified in a specific way when the pattern is imaged on a tilted surface. Mapping the imaged encoding or pattern and comparing it with the original encoding or mapping enables the tilt angle to be inferred based on the geometric relationship between the encoding / mapping modification and the surface tilt.

[0024] In general, the tilt angle of the sample surface may be determined by comparing the expected or reference geometry of the projected pattern with the modified geometry observed in the imaged pattern. This may involve analyzing the differences in one or more spatial relationships between pattern elements. In other words, based on corresponding pattern geometries in the projected and imaged patterns, the tilt angle of the sample can be calculated. In various examples, the dual-beam metrology system comprises a first charged-particle beam and a second charged-particle beam, originating from different charged-particle sources. The first and second particle beams are positioned at a fixed angle a relative to each other. For example, the first and second particle beam may be positioned at an angle a relative to each other when directed towards a sample, in particular a common position on the sample surface. But due to the relatively large distance between the particle sources and the sample and the relatively small dimension the sample surface, the angle may be regarded as general characteristic for the dualbeam device. The angle a may be defined between the general beam directions of the first and the second charged-particle beam when directed at the sample, wherein small directional deviations due to beam deflection due to scanning over the sample surface may be neglected. The first and second particle beam may be positioned at an angle a relative to each other with respect to their respective beam directions towards the sample surface, in particular a common incident location on the sample surface, which may, for example, comprise the first and / or second marking position, and / or any other position on the sample surface.

[0025] The first charged-particle beam may be used to process or modify the sample surface. For example, the first charged-particle beam may correspond to a Focused Ion Beam (FIB) particle beam, however other charged-particle beams could be used to modify the sample surface such that the modification can be detected in an imaging method based on the second charged-particle beam, such as, for example an electron beam.

[0026] The second charged-particle beam may be used to image the sample surface. For example, the second charged-particle beam may correspond to a Scanning Electron Microscope (SEM) electron beam, however other imaging modalities may be possible that rely on scanning the sample surface with a charged-particle beam.

[0027] In a step, a first marking may be applied at a first marking position on the sample surface using the first charged-particle beam. When applying the first marking the first particle beam may be arranged in a first beam position and / or beam orientation, also referred to as first beam pose or beam path, with regard to the sample surface and / or the device coordinate system. In an optional step, after marking the sample surface at the first marking position, the first charged-particle beam may be shifted to a second marking position on the sample surface. When applying a second marking, the first particle beam may be arranged in a second beam position and / or beam orientation, also referred to as second beam pose or beam path, with regard to the sample surface and / or the device coordinate system.

[0028] The first charged-particle beam may be shifted from the beam position at the first marking to a beam position at the second marking position. The first charged-particle beam when in the first beam position may be distanced from the first charged-particle beam when in the second beam position. That is to say, although the first charged-particle beam is sequentially arranged at the first and second beam positions, i.e., at the first and second marking for applying the respective marking, the first and second beam positions may be distanced from each other by the respective beam distance, e.g., dx. The first and / or second marking positions respectively beam positions may refer, for example, to a center position of the first respectively second markings. The first marking beam may be shifted with regard to the center positions, or in general to corresponding positions, which the first and second markings represent. In general, the inventive idea is based on the finding that by knowledge of the distance between the first charged-particle beam at the first and second markings, the first charged-particle beam may be relied upon as a pattern projector.

[0029] Therein, the first charged-particle beam may be shifted by the first beam distance dx. The first charged-particle beam when positioned at the second marking position may be distanced by dx from the first charged-particle beam when positioned at the first marking. The beam shift or distance may be characterized in a traversal plane to the beam direction, for example in the first and / or second marking positions. The first charged-particle beam may be shifted in a direction away from the first marking position in a perpendicular direction relative to the beam direction of the first charged-particle beam. Generally speaking, the first particle beam may be shifted along the sample surface to the second marking position on the sample surface.

[0030] It is to be understood that the first charged-particle beam at the first marking position and the first particle beam at the second marking position can be regarded as essentially parallel to each other for the purpose of the triangulation technique according to the present disclosure. Therefore, the distance between the first charged-particle beam at the first and second marking positions, should include any distance between the two beams near the sample surface. In general, the first charged-particle beams at the different positions can be distanced from each other by the respective beam distances. Therefore, the distance between the different beam positions, where the first charged-particle beams are arranged when applying the different respective markings, may be referred to as generally the beam distance, which does not substantially vary near the sample.

[0031] In particular, the first charged-particle beam, when positioned at the second marking position, can be offset, or distanced from the charged-particle beam at the first marking position by a first beam distance dx. In other words, the first particle beam may be shifted by the first beam distance dx in a traverse direction relative to the beam direction. The first beam distance may be in a traverse or perpendicular plane with respect to the first charged-particle beam. Thus, in a traverse plane, in other words perpendicular plane, to the beam direction of the first charged-particle beam in the first marking position and / or second marking position, the first charged-particle beam in the second marking position may be distanced from first charged-particle beam in the first marking position by the first beam distance dx. Similar considerations apply for any other markings, which may be applied by further shifting the first charged-particle beam, and / or for the distances between the second (imaging) charged-particle beams when directed to the different markings for imaging.

[0032] In a further step, a second marking may be applied at a second marking position on the sample surface using the first charged-particle beam.

[0033] Therefore, also the first and second markings, when projected onto a traverse, or perpendicular, plane with regard to the beam direction of the first charged-particle beam at the first or second marking, may be distanced from each other in the traverse plane by the beam distance dx. This is because the beams, or beam paths, of the first charged-particle beam when positioned at the first marking position as compared to when positioned at the second marking position on the sample surface are distanced, in other words offset, from each other by the first beam distance dx. In various examples, the first beam distance, and therefore also the first and / or second markings may lie in a plane defined by the first charged-particle beam and the second charged-particle beam, when positioned at the first marking position and / or the second marking position. This may allow particularly precise measurements, as the tilt angle determined in a direction with high measurement sensitivity. However, also any other directions are possible.

[0034] In various examples, the second charged-particle beam may be arranged substantially perpendicular, however with a small tilt angle, to the sample surface. This may allow particularly precise measurements, as the tilt angle determined in a direction with particularly high measurement sensitivity. However, also other directions are possible.

[0035] In various examples, when the first charged-particle beam is positioned at the second marking position, the first charged-particle beam, or a beam axis defined by the beam direction in the second marking position, beam may be distanced from the first marking position by a first beam distance dx. In various examples, when the first charged-particle beam is positioned at the first marking position, the first charged-particle beam, or a beam axis defined by the beam direction in the first marking position, may be distanced from the second marking position by a first beam distance dx.

[0036] The first charged-particle beam may be scanned over the sample surface. The first charged- particle beam may originate from a first beam source of a charged-particle beam sample processing module in direction towards the sample surface. The first charged-particle beam may be deflected by electrical fields in order to scan the sample surface, e.g., to reach the first and / or second marking positions. With regard to the beam distance, the beam may be regarded as represented by first charged-particle beam axes along the beam propagating directions, at the first and / or the second marking positions.

[0037] Therefore, due to the relatively large distance between the first charged-particle beam source to the sample surface, compared to the relatively small distance between the first and the second markings on the sample surface, the first charged-particle beam may be considered substantially parallel when positioned at the first marking position on the sample surface compared to the first charged-particle beam positioned at the second marking on the sample surface. In this regard the beam paths may be considered offset from each other by the first beam distance dx. For determining the sample tilt, the relatively small directional changes introduced by deflecting the first charged-particle beam to the fist respectively the second marking positions, may be neglected, and the beams may be considered parallel to each other. Similar considerations apply to the directions of the second charged-particle beam at the sample surface.

[0038] In other words, the first beam distance dx may be known in the direction of the first charged- particle beam, therefore a distance between the first marking position or the second marking position on the sample surface may be calculated. The first beam distance dx may be known as it can be adjusted as known in the art.

[0039] When using the beam distance as characterized above, the first and second markings are characterized by the beam distance as seen from the direction of the first charged-particle beam. The markings may be further characterized as seen from the direction the second charged-particle beam, for example by imaging the first and second markings using the second charged-particle beam, as will be described in the following.

[0040] In a further step, the second charged-particle beam may be scanned across the sample surface to acquire imaging data for an image of the sample surface including the first and second markings.

[0041] From the acquired imaging data, and / or the acquired image, a first imaged marking position of the first marking and a second imaged marking position are determined. The first and / or second imaged marking positions refer to the marking positions as seen from the angle of the imaging system, i.e., the second charged-particle beam. The first and second imaged marking positions may be determined in the image.

[0042] Based on the known beam distance dx, which characterizes the distance between the first and second markings as seen from the direction of the first (marking) charged-particle beam, the known angle a between the direction of the first (marking) charged-particle beam and the second (imaging) charged-particle beam, and the first and second markings as imaged in the direction of the second (imaging) charged-particle beam, in a further step, a first tilt angle of the sample surface may be determined. The first tilt angle may be determined in a machine coordinate system relative to the imaging beam along a first tilt direction between the first and second markings.

[0043] As the first and second markings are characterized from different directions, which include the known angle a in between them, these two characterizations may be used to determine a deviation, i.e., a first tilt angle, of the sample surface relative to the beam direction of the second charged-particle beam.

[0044] The first tilt angle may be determined based on a triangulation computation based on, or using, the angle a, the first beam distance dx, and the first and second imaged marking positions.

[0045] Thereby, an improved method is provided for determining sample tilt by using the first sample processing charged-particle beam (e.g. the FIB) as a pattern projector and the second imaging charged-particle beam (e.g. a SEM beam) in a known angle a to the first charged-particle beam as a triangulation sensor or camera, thereby enabling precise geometrical measurements on the sample surface using the actual processing and imaging hardware as measurement means. This approach leverages the existing hardware configuration of processing / marking modules and imaging modules of the dual-beam system, and therefore reduces systematical or inherent errors when using measured sample tilt for image correction, making it cost-effective and easy to implement in existing workflows. The provided techniques enable on-the-fly correction of sample tilt, e.g., during the slice-and-image process, thereby improving image quality and accurate 3D reconstruction.

[0046] In various examples, determining the first tilt angle may further comprise determining the first tilt angle by determining a first imaged marking distance in the image between the first and second imaged marking positions, and determining the first tilt angle using the first beam distance dx, the first imaged marking distance, and the angle a. Therein, the first tilt angle is calculated based on the known beam distance dx, the measured distance between the markings in the image, and the known angle a between the two charged-particle beams, using a triangulation.

[0047] In various examples, the first charged-particle beam may be used as a pattern projector for projecting a pattern comprising the first and second markings onto the sample surface. The second charged-particle beam may be used as a triangulation camera for measuring the first and second imaged marking positions corresponding to the first and second marking positions on the sample surface. In this configuration, the first tilt angle may be determined based on a triangulation computation using the beam distance dx, imaged marking distance, and the angle a as the triangulation angle between the first charged-particle beam and the second charged-particle beam.

[0048] Determining the sample first tilt angle may include computing at least one trigonometric formula using the angle a and / or the first tilt angle. Trigonometric formulas, such as one or more of the sine or cosine or tangent functions, or their arcsine, arccosine, and arctangent counterparts, can be used to calculate the tilt angle from the known parameters. For example, the tilt angle may be determined by the arctangent of the ratio of the measured distance to the product of the known beam distance and the cosine of the fixed angle.

[0049] The first charged-particle beam in the first marking position and the first charged-particle beam in the second marking position may be substantially parallel to each other. Similarly, the second charged-particle beam in the first marking position and the second charged-particle beam in the second marking position may be substantially parallel to each other.

[0050] Each one or both of these assumptions allow improved determination of the tilt angle using the first and second charged-particle beams, as the beam distance dx then corresponds, i.e. is substantially equal, to the distance between the first and second markings as seen from the direction of the first charged-particle beam, i.e. projected onto a perpendicular plane to the first charged-particle beam direction. Any deviations in the beam directions are negligible for determining sample tilt.

[0051] The second marking may he in a plane, which is defined by the first charged-particle beam in the first marking position and the second charged-particle beam when directed towards the first marking position. In other words, when the first charged-particle beam is moved from the first marking position to the second marking position, it is moved in plane with the angle a between the first and second charged-particle beams. This ensures that the markings and the two charged- particle beams are in the same plane, allowing for more accurate triangulation. The first charged-particle beam may originate from a Focused Ion Beam (FIB) module, and the second charged-particle beam may originate from a Scanning Electron Microscope (SEM) module, of the dual-beam metrology system.

[0052] Determining the first and second imaged marking positions in the image may include determining a center position of each of the first and second markings in the image. The center positions provide precise reference points for measuring the distance between the markings in the image. For tilt angle determination only the (exact) positions of the markings, not the form or any other dimension of the one of the markings is taken into account.

[0053] Before applying the first marking and / or the second marking, the method may include imaging the sample surface to acquire a reference image of the sample surface. A difference image may be generated based on the reference image and the image, wherein the difference image represents modifications of the sample surface due to the applied first and / or second markings. The first and / or second imaged marking positions of the first and second markings may be determined using the difference image. By comparing the reference image (before marking) with the image after marking, the positions of the markings can be clearly identified in the resulting difference image. Likewise, any difference between reference and imaged markings, for example a difference in the acquired imaging data may be used.

[0054] As will be appreciated, a similar approach may be used for determining a sample tilt angle in any other direction along the sample surface. For example, a second tilt angle in a second direction may be determined perpendicular to the first direction.

[0055] In various examples, the method for determining the second tilt angle may be based on determining two separate tilt angles in the first direction for two different point pairs, as will be described in the following.

[0056] Accordingly, the method may further include applying at least a third marking at a third marking position on the sample surface using the first charged-particle beam. When the first charged- particle beam is positioned at the third marking position, the first charged-particle beam may be distanced from the first marking position by a second beam distance dy in a perpendicular direction relative to the first beam separation direction dx. For the geometric arrangement of second beam distance dy the same considerations apply as for the first beam distance dx. In other words, the second beam distance dy may he in the traverse plane and perpendicular to the first beam distance dx.

[0057] In a step, the method may comprises applying, using the first charged-particle beam, a fourth marking at a fourth marking position on the sample surface, wherein the first charged-particle beam is in a fourth beam position.

[0058] When applying the fourth marking, the first charged-particle beam in the fourth beam position is distanced from the charged-particle beam in the third beam position by the first beam distance dx at the sample.

[0059] A third imaged marking position of the third marking may be determined from the imaging data or the image. A fourth imaged marking position of the fourth marking may be determined from the imaging data or the image.

[0060] The method may then determine a second tilt angle of the sample surface in a second tilt direction perpendicular to the first tilt direction between the first and third markings based on, or using, the angle a, the second beam distance dy, and the first, second, third, and fourth imaged marking positions. As described for the first tilt angle, the second tilt angle may be determined, based on a triangulation computation, and / or based on at least one trigonometric formula including the second tilt angle.

[0061] This allows for determining the sample tilt in different directions, specifically in two perpendicular directions, providing a more complete characterization of the sample surface orientation.

[0062] The determining of the second tilt angle may comprise determining a second imaged marking distance between the first and third imaged marking positions, and computing the second tilt angle based on at least one trigonometric formula using the second beam distance dy and the second imaged marking distance. Similar to the determination of the first tilt angle, the second tilt angle can be calculated using the known beam distance dy in the perpendicular direction and the measured distance between the first and third markings in the image.

[0063] The method may further comprise using the first and / or the second tilt angle for correcting at least one image acquired using the second charged-particle beam. The corrected image may be acquired additionally, and may comprise also a series of images, or it may comprise the image including the imaged markings. The tilt angles provide information about the sample surface orientation, which can be used to correct for perspective distortions in the acquired images, resulting in more accurate representations of the sample surface.

[0064] Therefore, the disclosed techniques do not rely on measuring a modification of a marking on the sample surface by the sample processing, which may be less precise due to the small marking dimensions, but on measuring precise positions of each marking, respectively a separation distance between the markings, which leads to a more precise measurement of sample tilt.

[0065] The method may be used in a 3D tomography process on the sample. The method may include performing a 3D tomography of the sample and reconstructing a 3D volume data set of the sample using the determined first and / or second tilt angle to correct for perspective distortion. By incorporating the tilt angle information into the 3D reconstruction process, the spatial accuracy of the resulting 3D volume data set can be improved, providing a more faithful representation of the sample's three-dimensional structure.

[0066] In various examples, the steps of the method for determining the first and / or second tilt angles may be repeatedly performed after a predetermined number of cycles of removing a layer from the sample surface and acquiring an image from the sample surface. This allows for updating the tilt angle information during the course of a multi-cycle material removal and imaging process, such as in a slice-and-view or serial sectioning procedure, ensuring that the tilt correction remains accurate throughout the process.

[0067] The first, second, and optionally third markings may be applied onto the sample surface in a marking region adjacent to a region of interest (ROI) to be imaged. By placing the markings close to, but outside of, the ROI, the tilt angles can be determined without interfering with the imaging of the ROI itself.

[0068] Alternatively, the first, second, and optionally third markings may be applied within a region of interest (ROI) to be imaged. This allows for direct determination of the tilt angles within the ROI, which may be advantageous if the sample surface orientation varies significantly across the sample.

[0069] The determining of the second tilt angle may comprise determining a second imaged marking distance between the first and third imaged marking positions, and computing at least one trigonometric formula using the second beam distance dy and the second imaged marking distance. Similar to the determination of the first tilt angle, the second tilt angle can be calculated using the known beam distance dy in the perpendicular direction and the measured distance between the first and third markings in the image.

[0070] The method may further comprise using the first and / or the second tilt angle for correcting at least one image acquired using the second charged-particle beam. The tilt angles provide information about the sample surface orientation, which can be used to correct for perspective distortions in the acquired images, resulting in more accurate representations of the sample surface.

[0071] The method may include performing a 3D tomography of the sample and reconstructing a 3D volume data set of the sample using the determined first and / or second tilt angle to correct for perspective distortion. By incorporating the tilt angle information into the 3D reconstruction process, the spatial accuracy of the resulting 3D volume data set can be improved, providing a more faithful representation of the sample's three-dimensional structure.

[0072] In an embodiment, the steps of the method for determining the first and / or second tilt angles may be repeatedly performed after a predetermined number of cycles of removing a layer from the sample surface and acquiring an image from the sample surface. This allows for updating the tilt angle information during the course of a multi-cycle material removal and imaging process, such as in a slice-and-view or serial sectioning procedure, ensuring that the tilt correction remains accurate throughout the process.

[0073] The first, second, and optionally third and fourth markings may be applied onto the sample surface in a marking region adjacent to a region of interest (ROI) to be imaged. By placing the markings close to, but outside of, the ROI, the tilt angles can be determined without interfering with the imaging of the ROI itself.

[0074] Alternatively, the first, second, and optionally third and fourth markings may be applied within a region of interest (ROI) to be imaged. This allows for direct determination of the tilt angles within the ROI, which may be advantageous if the sample surface orientation varies significantly across the sample.

[0075] The first and second beam distances dx and dy may be different from each other. This allows for greater flexibility in the marking process and may provide additional information and accuracy for determining the sample tilt angles in different directions, as may be needed for different (e.g., semiconductor) structure sizes in different directions.

[0076] The method may further comprise applying a plurality of further markings onto the sample surface positioned in a 2D pattern using the first charged-particle beam. A plurality of further tilt angles may be determined based on the further markings. The first and / or second tilt angle may be determined further based on a statistical analysis of the further tilt angles. By using a larger number of markings arranged in a two-dimensional pattern, the tilt angles can be determined with higher accuracy and precision, as the statistical analysis of the multiple tilt angle measurements can reduce the impact of individual measurement errors.

[0077] The spacings between the further markings in the 2D pattern may be equidistant in both directions of the 2D pattern on the sample surface. Alternatively, the spacings between the further markings in the 2D pattern may vary along at least one direction on the sample surface. The choice of equal or varying spacings may depend on the specific sample characteristics and the desired spatial resolution of the tilt angle determination, e.g., depending on structure sizes of the sample to be imaged. The two-dimensional pattern may, for example, be a regular array, such as a rectangular grid or a hexagonal grid, or an irregular array with varying spacings between the markings. The spacings between the additional markings in the two-dimensional pattern may be uniform in both dimensions, creating a pattern with equal distances between neighboring markings. Alternatively, the spacings between the additional markings may vary along one or both dimensions of the pattern. For example, the spacings may increase or decrease progressively along one direction, creating a gradient of marking densities. The variable spacings may be advantageous for adapting the pattern to specific sample characteristics or for obtaining tilt angle information with varying spatial resolution across the sample surface.

[0078] Determining the first and / or second tilt angle may further comprise a statistical analysis of the further tilt angles together with the spacings of the further markings. By considering both the tilt angles and the spacings between the markings, a more comprehensive analysis of the sample surface orientation can be performed, potentially leading to improved accuracy in the determined tilt angles.

[0079] These additional tilt angles provide a more comprehensive characterization of the sample surface orientation across the two-dimensional pattern. The first and / or second tilt angle may be refined or validated based on a statistical analysis of the additional tilt angles. For example, the statistical analysis may involve calculating an average, median, or mode of the additional tilt angles to obtain a representative value for the first and / or second tilt angle. The statistical analysis may also include determining a standard deviation or variance of the additional tilt angles to assess the variability or consistency of the sample surface orientation across the pattern.

[0080] The 2D pattern may, for example, comprise a line pattern or a crossing line pattern. A line pattern may be suitable for determining the tilt angle along a specific direction, while a crossing line pattern allows for determining the tilt angles in two perpendicular directions simultaneously.

[0081] The method may further comprise determining a curvature of the sample surface based on the further tilt angles. By analyzing the variation of the tilt angles across the 2D pattern, the local curvature of the sample surface can be estimated, providing additional information about the sample topography.

[0082] The method may include depositing a protective layer on the sample surface using a gas injection system (GIS) before applying the first and second markings. The first and second markings may then be applied onto the protective layer. The protective layer serves to prevent damage to the sample surface during the marking process and may also enhance the visibility of the markings in the acquired images.

[0083] The method may further comprise, before applying each respective marking of the first, second and optionally third markings, imaging the sample surface to acquire a respective individual reference image for each respective individual marking. Determining the imaged marking positions of the markings may then include, for each respective applied marking, generating an individual difference image by subtracting the respective individual reference image acquired before applying the respective marking from an image acquired after applying the marking (which could comprise, for example the reference image of the following marking, or the image acquired after applying the third marking). Each individual difference image represents changes on the sample surface due to the respective single individual marking only. By acquiring individual reference images for each individual marking before applying each marking and generating individual difference images for each individual marking, the method can more precisely identify the position of each marking, even in the presence of pre-existing surface features or irregularities caused by the marking processes of other markings. This approach enhances the robustness and reliability of the marking position determination, as it isolates the changes caused by each individual marking from the background surface structure or caused by other marking procedures.

[0084] The method may include analyzing the reference image acquired before applying each marking to identify one or more regions of the sample surface having a brightness satisfying a predetermined homogeneity threshold measure. A region may be selected from the one or more identified regions for applying the markings, and the markings may be applied within the selected region on the sample surface. By selecting a region with a homogeneous brightness for applying the markings, the method ensures that the markings are placed on a suitable area of the sample surface where they can be easily distinguished from the background. This selection process improves the visibility and contrast of the markings in the acquired images, facilitating accurate determination of their positions. The predetermined homogeneity threshold measure may be based on factors such as the brightness uniformity, gradient, or local variance within the region, and it can be adjusted depending on the specific sample characteristics and imaging conditions.

[0085] Therefore, determining the first and / or second imaged marking distance may be performed using a single image, and not comparing measurements from two or more images depicting the (possibly modified) markings. In other words, the second charged-particle beam may provide the necessary relative position information about the markings relative to each other, e.g. by determining positions of the markings from the same dataset, and not a positional shift or other dimensional change of one or both of the markings as determined e.g. at different time points or over two or more images. By using only the marking positions determined in a common image processing step, the method improves the tilt angle calculation process and reduces the overall measurement time.

[0086] A corresponding dual-beam metrology device and dual-beam metrology system may be configured to perform the method, or any combination of methods as disclosed. The device or system may comprise a charged-particle beam device, such as a charged-particle beam sample imaging module (e.g., scanning electron microscope module) and a charged-particle beam sample processing module (e.g., focused ion beam module), capable of generating and controlling the first and second charged-particle beams. The system may further include a sample stage for holding and positioning the sample, as well as a detector for acquiring images of the sample surface.

[0087] The system may further incorporate a controller comprising a processor and a memory for executing the method steps and storing the acquired data. The processor may be a general- purpose computer processor, or a specialized processor designed for image processing and analysis. The memory may include volatile memory, such as random-access memory (RAM), and non-volatile memory, such as read-only memory (ROM), flash memory, or a hard disk drive. The controller may be configured to cause the system to perform any method or combination of methods or methods steps, as describe in the present disclosure. In particular, in order to control the charged-particle beam device to apply the first and second markings onto the sample surface using the first charged-particle beam. It may also control the scanning of the second charged- particle beam across the sample surface to acquire the image including the first and second markings. The processor may further be programmed to determine the first and second imaged marking positions in the acquired image and calculate the first tilt angle of the sample surface based on the angle a, the first beam distance dx, and the first and second imaged marking positions.

[0088] The memory may store the acquired images, the determined marking positions, and the calculated tilt angles. It may also store the software instructions and algorithms necessary for performing the method steps, such as image processing routines, pattern recognition algorithms, and trigonometric calculations.

[0089] The dual-beam metrology device or system may include additional components, such as a user interface for inputting parameters and displaying results, a network interface for communicating with external devices or databases, and a storage device for long-term data storage.

[0090] The system may be configured to perform the method steps in an automated manner.

[0091] The dual-beam metrology device may be integrated into a larger metrology or inspection system, such as a semiconductor wafer inspection tool or a material characterization instrument. In such cases, the system may share common components, such as the sample stage, the vacuum chamber, or the control electronics, with other process modules of the larger metrology system.

[0092] By configuring the dual-beam metrology system to perform the disclosed methods, the system can accurately determine the sample tilt angles for correcting image distortions and enabling precise measurements of the sample surface.

[0093] In various examples, the computer-implemented method includes obtaining one or more images of a test volume on the wafer. The wafer is mounted on a sample stage of a dual-beam device. The one or more images are acquired using an imaging column all of the dual-beam device. The one or more images depict one or more cross-sections of the test volume of the wafer at one or more stage tilts of the sample states. The one or more images are obtained by slanted milling of the wafer using a milling column of the dual-beam device.

[0094] Under another aspect of the disclosure, the disclosed images and techniques can also be used for determining the angle a between the first and second charged-particle beams of the dual-beam metrology system, for example for calibration of the dual-beam metrology device. In various examples, this may be applied in dual-beam devices with variable angle a between the different charged-particle beams. In various examples, this may be advantageous, in order to determine the angle a precisely before performing any method or any combination of methods for determining sample tilt as described in the present disclosure.

[0095] A method is provided for determining or calibrating an angle a between a first charged-particle beam and a second charged-particle beam of a dual-beam device, wherein the first charged- particle beam is arranged in a first direction towards a sample, and the second charged-particle beam is arranged in a second direction towards the sample, wherein the first and second directions define the angle a relative to each other.

[0096] In an exemplary embodiment, the method comprises applying, using the first charged-particle beam, a first marking at a first marking position on a sample surface, wherein the first charged- particle beam is in a first beam position and the sample is in a first sample position. In a further step, the method further comprises scanning the second charged-particle beam across the sample surface to acquire a first image of the sample surface including the first marking while the sample is in the first sample position. The sample is then moved in the second direction, in particular in z-direction, by a distance Az to a second sample position. In a further step, a second marking is applied, using the first charged-particle beam, at a second marking position on the sample surface, wherein the first charged-particle beam is in the first beam position and the sample is in the second sample position. In a further step, the second charged-particle beam is scanned across the sample surface to acquire a second image of the sample surface including the first and second markings while the sample is in the second sample position. In a further step, the method comprises determining a first imaged marking position of the first marking in the first image and a second imaged marking position of the second marking in the second image. In a final step, the angle a is determined based on the first and second imaged marking positions.

[0097] The sample may be moved translationally into the second sample positions. The sample may be moved only in or at least with a component in in the second direction, particularly in z-direction, for example using a movable stage.

[0098] Optionally, the position of the first marking may also be determined in the second image. This additional measurement can be used to correct for any translational movements of the sample in the plane perpendicular to the second direction / z-direction, i.e., in the x or y directions. By comparing the position of the first marking in the first image with its position in the second image, a positional difference of shift of the sample can be determined and accounted for in the determination of the angle a.

[0099] To account for any tilt of the stage or sample surface relative to the beam axes, the measurement may be repeated after rotating the sample by a predetermined angle, particularly by 180 degrees, around an axis perpendicular to the sample surface.

[0100] The sample may be rotated by 180 degrees about the first marking position, which serves as the center of rotation. This rotation is performed after the first measurement of the angle a, which involves applying the first and second markings at the first and second sample positions, respectively, and acquiring the corresponding images.

[0101] After the rotation, the measurement process is repeated in a second angle determination process. The true angle a between the first and second charged-particle beams can then be calculated by taking the average of the two measurements. This averaging helps to eliminate the effect of any tilt of the stage or sample surface, as the tilt's influence on the angle measurement is opposite for the two measurements taken 180 degrees apart.

[0102] A computer program or a computer-program product and a computer-readable storage medium is provided that includes program code. The program code can be loaded and executed by at least one processor. Upon loading and executing the program code, the at least one processor causes a dual-beam device to perform any method, or any combination of methods as disclosed. It is to be understood that the features mentioned above and features yet to be explained below can be used not only in the respective combinations indicated, but also in other combinations or in isolation, without departing from the scope of the present disclosure. In particular, the features mentioned above and those yet to be explained below may be used not only in the respective combinations indicated, but also in other combinations or in isolation without departing from the scope of the disclosure.

[0103] BRIEF DESCRIPTION OF THE DRAWINGS

[0104] These and other objects of the invention will be appreciated and understood by those skilled in the art from the detailed description of the preferred embodiments and the following drawings in which like reference numerals refer to like elements.

[0105] Fig. 1 schematically illustrates a slice-and-image process using a dual-beam device, according to various examples.

[0106] Fig. 2 schematically illustrates a 3D tomography image reconstruction for multiple 2D images obtained using the slides-and-image process according to the example of Fig. 1.

[0107] Fig. 3 is a detailed view of a cross-section obtained by the slides-and-image process as shown in Fig. 1 for a wafer that includes a 3D memory semiconductor device structure according to various examples.

[0108] Fig. 4 schematically illustrates a dual-beam device configured for executing a slice-and-image process, according to various examples.

[0109] Fig. 5 schematically illustrates multiple degrees of freedom in positioning a wafer stage, according to various examples.

[0110] Fig. 6 schematically illustrates wafer bow in a dual-beam device, according to various examples. Fig. 7 schematically illustrates a tilt angle of a sample surface of the dual-beam device of Fig. 6 due to wafer bow, according to various examples.

[0111] Fig. 8 schematically illustrates a triangulation technique, which can be applied in the dual-beam devices of Figs. 1-7, according to various examples.

[0112] Fig. 9 schematically illustrates a further triangulation technique, which can be applied in the dualbeam devices of Figs. 1-7, according to various examples.

[0113] Fig. 10 schematically illustrates a triangulation technique based on applied markings, according to various examples.

[0114] Fig. 11 is a flowchart of a method for determining a tilt angle of a sample, according to various examples.

[0115] Fig. 12 schematically illustrates a triangulation technique for determining the angle a, that can be applied in the dual-beam devices of Figs. 1-7, according to various examples.

[0116] FIG. 13 is a flowchart of a method for determining an angle a between a first and a second charged-particle beams of a dual-beam device, according to various examples.

[0117] DETAILED DESCRIPTION

[0118] In the following, embodiments of the invention will be described in detail with reference to the accompanying drawings. It should be understood that the following description of embodiments is not to be taken in a limiting sense. The scope of the invention is not intended to be limited by the embodiments described hereinafter or by the drawings, which are taken to be illustrative examples of the general inventive concept. The features of the various embodiments may be combined with each other, unless specifically noted otherwise.

[0119] Some examples of the present disclosure generally provide for a plurality of circuits or other electrical devices. All references to the circuits and other electrical devices and the functionality provided by each are not intended to be limited to encompassing only what is illustrated and described herein. While certain labels may be assigned to the various circuits or other electrical devices disclosed, such labels are not intended to limit the scope of operation for the circuits and the other electrical devices. Such circuits and other electrical devices may be combined with each other and / or separated in any manner based on the type of electrical implementation that is desired. It is recognized that any circuit or other electrical device disclosed herein may include any number of microcontrollers, a graphics processor unit (GPU), integrated circuits, memory devices (e.g., FLASH, random access memory (RAM), read only memory (ROM), electrically programmable read only memory (EPROM), electrically erasable programmable read only memory (EEPROM), or other suitable variants thereof, and software which co-act with one another to perform operation(s) disclosed herein. In addition, any one or more of the electrical devices may be configured to execute a program code that is embodied in a non-transitory computer readable medium programmed to perform any number of the functions as disclosed.

[0120] The drawings are to be regarded as being schematic representations and elements illustrated in the drawings are not necessarily shown to scale. Rather, the various elements are represented such that their function and general purpose become apparent to a person skilled in the art. Any connection or coupling between functional blocks, devices, elements, or other physical or functional units shown in the drawings or described herein may also be implemented by an indirect connection or coupling.

[0121] Various techniques disclosed relate to tomographic imaging of semiconductor wafers. For instance, 3-D tomographic data of semiconductor structures on nm scale can be obtained through the so-called slice-and-image process that can be implemented using a dual-beam device (DBD). In various examples, two charged particle systems can be arranged at an angle (column offset angle). The two particle systems might be oriented at a column offset angle between 45° and 90°. A first charged particle system may define a milling column. The milling column can be a focused ion beam (FIB) system, using for example Gallium (Ga) ions. The FIB of Ga ions may be used to cut off slices of a test volume of the wafer, slice-by-slice. Thereby, images depicting cross-sections of the wafer are obtained at different milling depths, using the imaging column.

[0122] A second particle optical system may define an imaging column. The imaging column can, for example, be implemented by a scanning electron microscope (SEM) or a helium-ion microscope (HIM). The relevant background of the DBD and operation of the DBD, including the slice-and- imaging approach will be explained in connection with FIGs. 1 to 4.

[0123] FIGS. 1 to 5 schematically illustrate dual-beam devices or systems, with which a method for determining sample tilt, as described in the present disclosure, can be performed.

[0124] Fig. 1 schematically illustrates a slice-and-image process using a dual-beam device, according to various examples.

[0125] As can be seen in Fig. 1, a slice-and-image process can be performed using a DBD 500: using a milling column 550, with a focused ion beam 551 in z-direction, and scanning in y-z-plane, a thin layer of a cross-section through a test volume 510 of a wafer 511 is removed to reveal a new front cross-section 552 as a cross section of the test volume 510 (after imaging, this can be repeated to reveal further cross-sections 553, 554).

[0126] The imaging column 540 - e.g., a SEM - is used for scanning imaging of the cross-sections associated with the cross-sections 552, 553, 554.

[0127] The column offset angle 597 is only schematically illustrated in this example. Various values are possible for the column offset angle 597. The column offset angle may also describe the angle between the imaging charged-particle beam 544 and the milling charged-particle beam 551 when directed at the sample, for example a common position on the wafer 511 surface.

[0128] In the illustrated example, the SEM optical axis 542 is shown, and the image is generated by scanning the electron beam 544 along scanning imaging lines 546. After performing the raster scan of the cross-section 552, a two-dimensional (2-D) image 1001 is formed.

[0129] By repetition of this milling and imaging, further images 1002 and 1003 are obtained, at different milling depths 599. The images 1001, 1003, 1004 thus correspond to image slices obtained at different milling depths 599 of the test volume 510. The images 1001, 1002, 1003 can be labeled cross-sectional image slices. The depth offset between two subsequent milling depths can be between Inm and some tens of nm, for example 30nm. From the sequence of these 2-D images 1001-1103, a 3D tomographic image of the test volume, depicting the semiconductor structures 1010 can be reconstructed.

[0130] In general, various kinds and types of samples, e.g., semiconductor structures can be imaged. For instance, a three-dimensional (3-D) memory chip as shown in Fig. 2.

[0131] Fig. 2 is a 3D tomography image reconstruction for multiple 2D images obtained using the slides- and-image process according to the example of Fig. 1.

[0132] As can be seen in Fig. 2, a 3-D tomographic image 1030 is reconstructed from multiple 2-D images. The 3-D tomographic image 1030 depicts a 3-D memory device. The 3-D tomographic image 1030 has been acquired using a DBD such as a FIB-SEM DBD or a FIB-HIM DBD, and using a slice-and-image process, as explained in connection with Fig. 1.

[0133] If the milling column is arranged such that the slices are oriented perpendicular to the pillar / channel axes of the 3-D memory device, wherein other angles towards the sample surface would be possible. Thus, the slices can be “photographed” using the imaging column.

[0134] In other scenarios, it is possible that the optical axis 542 of the imaging column 540 is not oriented perpendicular to the imaged cross-sections. Such scenario is shown in Fig. 3.

[0135] Fig. 3 is a detail view of a cross-section obtained by the slides-and-image process of Fig. 1 for a wafer that includes a 3D memory semiconductor device structure according to various examples.

[0136] Fig. 3 illustrates slice-and-imaging approach for a 3-D memory device. The wafer top surface 511 A of the test volume 510 of the wafer 511 is milled with a FIB beam 551 at a milling angle 598 below 10° to the wafer surface, for example at milling angle of 8°, but other milling angles below 45° are possible as well, for example 40°, or 36°. A small milling angle 598 of between 8° and 45° may be preferred since is allows better imaging with high resolution. Such non-90° milling angles 598 are referred to as slanted milling. Fig. 3 illustrates multiple layers 1012 of the 3-D memory device. The multiple layers 1012 can implement word lines of the memory write / read process. Also illustrated are the semiconductor structures 1011.

[0137] Also illustrated in Fig. 3 is a cutting edge 552A of the cross-section 552. The cutting edge 552 is the edge between the wafer top surface 511A and the cross-section 552. The cutting edge 552A can be visible in the 2-D images.

[0138] Fig. 4 schematically illustrates a dual-beam device configured for executing a slice-and-image process, according to various examples.

[0139] Further aspects of a sample processing module with a first charged-particle beam and an imaging module with a second charged-particle beam are explained with respect to the DBD 500. The DBD 500 includes a sample stage 515 (or wafer stage) and a wafer 511 - which could be an entire circular wafer or only a small cut-out, a so-called coupon or die - is mounted on the sample stage 515. A test volume 510 is aligned at the intersection point 543 of the imaging column 540 and the milling column 550.

[0140] The sample stage 515 defines an X-Y-plane. The Z-axis is oriented perpendicular to the X-Y- plane. The X-axis, Y-axis and Z-axis are defined in a machine coordinate system of the DBD 500.

[0141] The optical axis 548 of the milling column 550 and the optical axis 542 of the imaging column 540 enclose the imaging offset angle 597, which can also be referred to as the angle between the two charged-particle beams of the imaging and the milling columns. In this examples, the optical axis 542 forms an angle 602 with the z-axis, which defined to be perpendicular to the wafer plane. In other examples, as described herein, the optical axis 542 may be perpendicular to the wafer plane.

[0142] At the intersection point 543, the wafer top surface 511 A is arranged at angle 601 to the optical axis 548 of the milling column 550. This milling angle 601 depends on a wafer tilt 120. With the FIB 551 impinging under milling angle 601 onto the top surface of the wafer 511, slanted crosssections are milled into the wafer by ion beam milling at the test volume 510.

[0143] The wafer tilt 120 between the wafer 511 and the sample stage 515 is illustrated. The wafer tilt 120 corresponds to an angle between the wafer top surface 511 A and the X-Y-plane; as such, it can have two components, corresponding to the X-axis and the Y-axis, respectively. The wafer tilt 120 results in an inclination of the wafer normal (perpendicular to the wafer top surface 511 A) with respect to the Z-axis. The wafer tilt 120 can occur due to imperfect mounting of the wafer 111 on the sample stage 515. For instance, it would be possible that small dust particles are present on the bottom surface of the wafer 511 and / or on the top surface of the wafer stage 515. This can cause the wafer till 120. Furthermore, sometimes, the wafer 111 is attached to the top surface of the wafer stage 515 using adhesive pads. These adhesive pads can have nonplanar surfaces. According to various examples, it is possible to determine the wafer tilt 120 using the DBD 550, for example, based on applying markings onto the sample surface using the milling column and measuring the markings by the imaging column, as described throughout the present disclosure.

[0144] During imaging, a beam of charged particles is scanned by a scanning unit of the imaging column 540 along a scan path over a cross-section surface of the wafer, and secondary particles as well as scattered particles are generated. A particle detector 517 collects at least some of the secondary particles and scattered particles and communicates the particle count with a control unit 519. The control unit 519 is in control of the imaging column 540 and the milling column 550 and connected to a control unit 516 to control the position of the wafer mounted on the wafer support table via the wafer stage 515. The wafer stage 515 can often be displaced in X, Y, and Z- direction, and can be rotated around X and / or Y direction, as will be further explain in connection with Fig. 5.

[0145] Fig. 5 schematically illustrates multiple degrees of freedom in positioning a wafer stage 515, according to various examples.

[0146] The stage 515 can usually be rotated around a given axis by a precise angle. That is, if the initial tilt of the stage around a controllable axis is 00, it can be rotated to a position corresponding to the tilt of (0O+ A0), where A0 can be set very precisely. It is, however, challenging to determine the zero-point 0Oof the stage tilt with a sufficient accuracy. In some cases, there can be inaccuracies in 0O, e.g., while nominally 0O:= 0, defacto 0O#= 0. These inaccuracies can lead to the stage being tilted against the X-Y-plane. Inaccuracies of the zero point of the stage tilt, thus, are considered herein to add to the wafer tilt.

[0147] A control unit 502 controls the overall acquisition protocol, stage positioning, triggers acquisition, reads out images, reconstructs 3-D tomographic images based on the machine coordinate system of the X-Y-Z-axes, and performs further post-processing. In some cases, the control unit 502 can determine the wafer tilt 120. The wafer tilt 120 can be taken into account (e.g., to rotate the machine coordinate system) when reconstructing the 3-D tomographic images or compensated for by changing the stage tilt of the wafer stage.

[0148] For instance, the control unit 502 could be implemented as a general-purpose processor that can load and execute program code from a non-volatile memory 502A.

[0149] Above, background aspects with respect to a slice-and-image process have been disclosed. The techniques described herein are based on the finding that sometimes the wafer tilt can affect the accuracy of quantitative analysis of semiconductor device structures, as will be further explained in the following FIGS. 6 to 8. Accordingly, techniques will be disclosed that facilitate to determine the wafer tilt 120 so that, subsequently, an accurate analysis of the semiconductor device structures is facilitated.

[0150] In the process of semiconductor tomography, a scanning electron microscope (SEM) is used to capture a multitude of images of a region of interest (ROI) on a wafer. The ROI is typically 5 pm x 5pm in size. Typically, a resolution of 5000 x 5000 pixels is used, with a pixel spacing of 1 nm. Higher resolutions and larger images are also possible.

[0151] Between the individual images, a focused ion beam (FIB) is used to remove a thin layer of a few nanometers from the surface. The focused ion beam is positioned at a defined angle to the optical axis of the SEM. Good results are achieved, for example, with an angle of 65° between the FIB and SEM, which corresponds to an angle of 25° between the FIB and the wafer surface (FIB- SEM angle a). This is known as a "slice and scan" method, an iterative process of removing a thin surface layer with the FIB followed by imaging. This allows for the collection of several thousand images.

[0152] All images are registered in a common coordinate system, and with the known layer thickness between the individual images, a high-resolution volumetric dataset can be calculated. It is crucial for the reconstruction of the 3D volumetric dataset from the individual images that the wafer surface is as perpendicular (90°) to the SEM beam as possible, or that the angle of the wafer surface to the SEM beam is precisely known so it can be considered during the reconstruction process. This ideal alignment is not guaranteed due to equipment limitations and the inherent unevenness of the wafer itself.

[0153] The wafer is positioned on a so-called wafer stage under the SEM. The wafer stage is a multi-axis mechanical positioning system consisting of two typically perpendicular horizontal axes (x and y), a vertical axis (z), and a vertical rotation axis (T). This setup allows any point on the wafer to be observed with the SEM at a defined working distance. The rotation axis (T) also allows the cutting direction of the FIB to be precisely set by rotating the wafer into the desired position.

[0154] The wafer itself is not perfectly flat but bowed. For 300 mm wafers, the so-called wafer bow can be up to 1 mm. An electrostatic chuck (E-chuck) can hold the wafer in place with electrical forces and flatten it to a large extent. However, due to the unevenness of the E-chuck and the residual bow of the wafer, it is not possible to completely flatten the wafer on the chuck. This means that the wafer surface exhibits different slopes (wafer slope) relative to the SEM beam at various points on the wafer.

[0155] Due to axis non-linearities, tolerances in the axis bearings, and the limited accuracy of the stage alignment to the SEM, additional tilts of the chuck relative to the SEM beam can occur depending on the positioning of the different axes. Additionally, the SEM is not mounted perfectly at a 90° angle to a reference plane, which contributes further to the overall tilt. The resulting tilt angle Q, caused by the wafer unevenness, chuck, and SEM tilt, means the angle between the SEM beam and the wafer surface is not exactly 90°. Fig. 6 schematically illustrates wafer bow in a dual-beam device, according to various examples.

[0156] As can be seen in Fig. 6, the schematic diagram of a dual-beam device configuration for examining a wafer comprises an imaging Scanning Electron Microscope (SEM) column and a milling Focused Ion Beam (FIB) column. The wafer is positioned on an electrostatic chuck (E- chuck), which holds the wafer in place during processing. Specifically, Fig. 6 demonstrates the alignment of the wafer under SEM / FIB with exaggerated illustration of wafer unevenness and chuck tilt.

[0157] The dual-beam device comprises two main columns, FIB and SEM, e.g., two main process modules, but further process modules can be added to the system, such as, for example a Gas Injection System (GIS). The FIB column sends a first charged-particle beam towards the sample surface. The SEM column sends a second charged-particle beam to the sample surface. The main axes, or longitudinal axis, or beam axes, of the FIB and SEM column, are arranged in an angle a relative to each other at the sample surface.

[0158] The imaging column is used for imaging the wafer surface and cross-sections, wherein the SEM emits a second charged-particle beam, used to scan the wafer surface to capture high-resolution images.

[0159] The milling column emits the first charged-particle beam that is used for milling or cutting the wafer to create cross-sections for imaging. The SEM and FIB columns are positioned at a fixed angle a, also referred to as the column offset angle, which is important for ensuring the beams converge accurately on the same area of the wafer for precise milling and imaging.

[0160] In the context of semiconductor tomography, a method called "Slice and Scan" is used. This involves alternately milling a thin layer of the wafer surface with the first charged-particle beam and then imaging the new surface with the second charged-particle beam. This iterative process allows the collection of multiple images that can be reconstructed into a high-resolution volumetric dataset. Good results are typically achieved with an angle a of 65° between the FIB and SEM, corresponding to a 25° angle between the FIB and the wafer surface. The wafer is depicted slightly tilted on the E-chuck, wherein a tilt adjustment or tilt correction may be necessary for optimal processing. In this example, the wafer is not perfectly flat; it has a bow that, for example, can be up to 1 mm for 300 mm wafers. The E-chuck helps flatten the wafer, but due to inherent inaccuracies and the wafer bow, the wafer surface can still exhibit a tilt or slope. This tilt, along with inaccuracies in the positioning system and the non-perfect alignment of the SEM column, contributes to a resulting tilt angle Q. This means that, in the example of Fig. 6, the angle between the SEM beam 544 and the wafer surface is not exactly 90°, as will be further explained in Fig. 7.

[0161] The angle measurement between the SEM beam 544 and the wafer surface in the region of interest can be performed using patterns applied, i.e., inscribed, on the wafer surface by the FIB. The FIB acts as a pattern projector at a triangulation angle a relative to the SEM beam, which functions like a triangulation camera. By measuring the distances between these inscribed points in the SEM image, the tilt angle of the wafer surface can be precisely determined, as will be further explained in the following.

[0162] Fig. 7 schematically illustrates the tilt angle of a sample surface of the dual-beam device of Fig. 6 due to wafer bow, according to various examples.

[0163] As can be seen in Fig. 7, the geometric relationship between the SEM beam and the wafer surface in a dual-beam system as shown in Fig. 6 is further illustrated in detail.

[0164] The SEM beam is at an angle that is not exactly 90° to the wafer surface. This deviation from the ideal 90° alignment by a tilt angle <F is due to several factors such as the inherent bow of the wafer, imperfections in the mounting of the wafer, and misalignments in the equipment.

[0165] The angle between the SEM beam and the wafer surface may be referred to as a tilt angle of the sample, or sample surface. The angle is shown with reference to a perpendicular direction to the beam direction of the second charged-particle beam of the imaging column. However it is clear that that tilt angle can also be reference to other axes, such as a surface normal to the sample surface at the measurement point, and / or the beam direction of the second (i.e., SEM) charged- particle beam. In general, the orientation of the sample surface with regard to the measurement system, or machine coordinate system, particularly with regard to the FIB and SEM charged- particle beams, can be determined.

[0166] This tilt is to be accurately measured and accounted for in order to ensure precise imaging and analysis, particularly in processes like 3D tomography where accurate reconstruction of the volumetric data depends on knowing the exact orientation of the wafer surface relative to the SEM beam.

[0167] Fig. 8 schematically illustrates a triangulation technique, which can be applied in the dual-beam devices of Figs. 1-7, according to various examples.

[0168] Fig. 8 shows the principle of triangulation with a pattern projector and an optical camera. It is to be understood, that the principle can be applied to any kind of pattern projector, such as for example a FIB module (cf. first charged-particle beam) modifying a sample surface for projecting the pattern, making the pattern visible for the camera. Also the principle can be applied to any sensor, which can detect the projected pattern, such as for example a SEM imaging module (cf. second charged-particle beam) for imaging the applied markings on the sample surface by the FIB module.

[0169] In Fig. 8, the sample is depicted arranged in a horizontal orientation in a first sample position 1, however it is to be understood that the principle can be applied to any orientation of the sample. From the sample position 1 the sample is translationally (e.g., not rotationally) moved by a distance Az in vertical direction to the horizontal second sample position 2. The sample shift in z- direction causes a change in image position of the projected point (marking) on the sample surface, as will be explained in the following.

[0170] The pattern projector is arranged in a fixed known triangulation angle a to the optical axis of the camera. The camera comprises an image sensor arranged in an essentially horizontal orientation to the optical beam path. The camera further comprises a lens, which optically deflects incoming optical beams from the sample onto the image sensor. In the first sample position, the pattern projector projects a pattern, comprising at least one marking, e.g., a point 800, onto a first marking position 810 on the sample surface. When the sample is positioned in the first sample position, the at least one marking is imaged by the lens onto the image sensor to a first image position 811 on the image sensor.

[0171] In the second sample position, the sample surface is moved vertically by the distance Az to a lower position. Due to the moved sample surface the pattern projector projects the pattern into the same direction, along the pattern projector beam direction, onto the sample surface, wherein the marketing is applied to the sample surface at a second marking position 820. The first marking position 810 and the second marking position 820 are horizontally distanced from each other.

[0172] When the lens images the second marking position onto the image sensor, the second imaged marking position 821 on the image sensor is shifted relative to the first image marking position 811 by a distance Ax on the image sensor, i.e., in the image plane.

[0173] The triangulation principle, as illustrated in Fig. 8, can be used as a method to determine the vertical displacement of a sample by analyzing the horizontal imaged shift of a projected pattern on the sample surface. This principle can be applied using various pattern projectors, such as for example a focused ion beam (FIB) module or a laser, and different sensors, like an optical camera or a scanning electron microscope (SEM) imaging module.

[0174] When the sample is vertically displaced, either raised or lowered, the position of the projected point on the sample surface shifts horizontally due to the triangulation angle. The sensor captures the new position of the marking on its image plane, and the horizontal shift of the marking relative to its initial position is measured.

[0175] By knowing the triangulation angle and measuring the horizontal shift of the marking on the sensor's image plane, it is possible to calculate the vertical displacement of the sample. This calculation is based on the geometric relationship between the triangulation angle, the horizontal shift, and the vertical displacement. Conversely, if the vertical displacement of the sample is known, along with either the triangulation angle or the horizontal shift of the marking, the remaining parameter can be determined using the same geometric relationship.

[0176] Fig. 9 schematically illustrates a further triangulation technique, which can be applied in the dualbeam devices of Figs. 1-7, according to various examples.

[0177] As can be seen in Fig. 9, the triangulation setup is the same as in Fig. 8, with the pattern projector arranged at a fixed triangulation angle a relative to the optical axis of the camera. The camera comprises an image sensor and a lens that optically deflects incoming light beams from the sample surface onto the image sensor.

[0178] However, instead of shifting the sample vertically by a distance Az, the sample is tilted by an angle <5. The sample tilt causes the projected pattern to appear shifted on the sample surface, as the surface orientation has changed relative to the pattern projector.

[0179] When the sample is in its original position (Sample Position 1), the pattern projector projects a pattern or marking, e.g., point 900, onto the sample surface, and the camera captures the image of the pattern at an imaged marking position 911. The lens focuses the light beams from the pattern onto the image sensor at imaged marking position 911.

[0180] As the sample is rotated to the tilted sample position, the surface orientation changes, causing the projected pattern to shift on the sample surface to shifted second marking position 920. The camera captures the new image of the shifted pattern at imaged marking position 921, wherein the lens focuses the light beams onto the image sensor at the new imaged position, shifted by a distance Ax relative to the first imaged marking position 911.

[0181] The shift Ax of the pattern on the image sensor is related to the tilt angle 0 of the sample, the triangulation angle a, and the distance between the sample and the camera. As described in the disclosed techniques, by measuring the shift Ax and using the given triangulation angle a and the camera-sample distance, the tilt angle 0 can be determined using trigonometric relationships. Fig. 10 schematically illustrates a triangulation technique based on applied markings in a dualbeam device as described for the previous figures, according to various examples.

[0182] As illustrated in Fig. 10, a tilt angle of the curved wafer surface relative to the SEM beam can be measured using a pattern of points inscribed by the FIB.

[0183] The FIB acts as a pattern projector, inscribing a series of points, or markings, onto the wafer surface at a known beam distances, also referred to as beam distances, between the marking FIB beams. These points are imaged and detected by the SEM module, where the imaged marking positions can be used with the known beam distances and fixed angle a to determine sample tilt.

[0184] The wafer surface has a curved surface due to wafer bow as described above. The inscribed points are arranged in a grid along this curved surface. The markings include a first marking 10, a second marking 20, a third marking 30, and a fourth marking 40. The markings in this example are depicted as points 10, 20, 30, 40.

[0185] In a first sample tilt direction, which may be referred to as x-direction, the imaged distance between the first marking 10 and the second marking 20 is Axl. The distance Axl between the first marking 10 and second marking 20 may be determined by imaging the markings using the SEM module from an essentially perpendicular direction from above (z-direction).

[0186] In the first sample tilt direction, the imaged distance between the third marking 30 and the fourth marking 40 is Ax2.

[0187] In a second sample tilt direction, which may be referred to as y-direction, perpendicular to the x- direction, the imaged distance between the first marking 10 and the third marking 30 is Ay.

[0188] The positions of the markings 10, 20, 30 are imaged by the second charged-particle beam of the imaging system (SEM module) of the previous figures (not depicted in Fig. 10), and accordingly the distances Ax and Ay correspond to imaged distances between the markings 10, 20, 30. Accordingly, the distances between the inscribed points in the x and y directions (Ax and Ay) correspond to distances as determined by the SEM module, in a direction substantially perpendicular to the sample surface, by measuring in the SEM image.

[0189] The beam distances dxl, dx2 and dy are depicted as the offset between the beam paths (shown in dotted lines) of the first charged-particle beam of the FIB module, when positioned to apply the markings at their respective marking positions. The beam distance dxl refers to the first and second markings 10, and 20. The beam distance dx2 refers to the third and fourth markings 30, and 40. The beam distance dy relates to the first and third markings 10, 30.

[0190] Accordingly, Fig. 10 illustrates that by shifting the marking beam by a predetermined, or known, beam distance relative to the previous beam path, each of the markings can be applied with known beam distance from a previous marking. The beam distances can be applied to the first charged-particle beam as known in the field. For example, when looking along the beam direction of the first charged-particle beam when located at the fist marking 10, the beam will be shifted such that in the second marking 20 position on the sample surface, a distance is applied in a traverse plane between the beam paths that equals dxl. Similar considerations apply to any other markings.

[0191] The first beam distance dxl and the imaged marking distance Axl between the first and second markings 10, 20 correspond to a first direction and a first tilt angle along the first direction. The second beam distance dy and the imaged marking distance between the first and third markings 10, 30 correspond to a second direction, perpendicular to the first direction, and a second tilt angle along the second direction.

[0192] This demonstrates how, in the context of a FIB-SEM arrangement, the Focused Ion Beam takes on the role of a pattern projector at a triangulation angle a relative to the SEM beam. In this configuration, the SEM can be considered as a triangulation camera. The angle between the FIB and SEM is known due to the geometric arrangement and potentially through calibration. This camera-projector unit enables geometric measurements on the wafer surface. In particular, the tilt angle of the wafer relative to the SEM beam can be measured precisely. The angle measurement is performed by writing at least two points on the wafer surface using the FIB. The distance between the points on the wafer is determined through triangulation calculations using the beam distance dx (the FIB beams can be considered approximately parallel), the triangulation angle a, and the tilt angle of the wafer surface. The FIB beams can be precisely adjusted to a specific value over a wide range through electrical beam deflection.

[0193] The formula for the point distance Ax is: dx

[0194] Ax = cos(<5) » - — - - cos (<5 — a)

[0195] For small angles <5, cos(<5) can be approximated as 1, thus: dx dx = - — - - cos (<5 — a)

[0196] When the two written points are projected in the x / z plane, the tilt angle <5 of the wafer surface around the y-axis can be calculated with a known FIB beam distance dx, if the point distance Ax is measured with the SEM:

[0197] <5 = arccos

[0198] The tilt angle of the wafer surface around the x-axis can be calculated from two inscribed point pairs by first calculating the tilt angles <51 and <52 around the y-axis separately for each point pair 10,20 and 30,40, and then calculating the tilt angle around the x-axis according to the following formula:

[0199] Therein, dy is the beam distance of the FIB beams in the y-direction. In general, it can be sufficient to write one pair of points for each direction (x and y) and measure the resulting distance between the points in the SEM image. However, a large number of points can also be written, and the average value, or any other statistical deprival, of individual angle measurements can be used to determine the mean tilt angle of the sample surface. This makes the determination of the tilt angle more reliable.

[0200] With a sufficient number of individual points, the topography of the wafer surface can be measured, and local curvatures in the imaged wafer region can also be determined.

[0201] In other words, the tilt angle of the sample surface relative to the second charged-particle beam in a first direction, can be determined based on a trigonometric relationship between: (i) the known angle between the first and second charged-particle beams,

[0202] (ii) the known beam distance between the first and second markings as applied by the first charged-particle beam, and (iii) the measured distance between the first and second markings in the image acquired by the second charged-particle beam.

[0203] Therein, the trigonometric relationships account for, i.e., represent, the geometric arrangement of the dual-beam device and the sample surface orientation.

[0204] Therefore, in general, using triangulation under the known angle a, and with the known beam distance dx, the imaged positions of the markings 10 and 20 in the SEM image can be used to calculate the first tilt angle along Axl .

[0205] In other words, using a triangulation consideration, the tilt angle of the wafer surface can be determined by comparing the measured point distances with the known separation of the FIB beams in the triangulation angle a.

[0206] This method allows for accurate measurement of the wafer tilt, ensuring precise alignment of the SEM and FIB for high-resolution imaging and reconstruction of the 3D volumetric dataset. The pattern of points can be adjusted for density and distribution to enhance measurement accuracy, and the tilt can be measured at various points across the wafer surface to account for local variations in curvature. Similarly a second tilt angle along Ay may be determined. The second tilt angle can, for example, be determined in a second direction perpendicular to the first direction based on the beam distance dy and an imaged marking Ay distance between the first and third markings 10, 30.

[0207] Fig. 11 is a flowchart of a method for determining a tilt angle of a sample, according to various examples.

[0208] The method starts in step S10. In step S20, using the first charged-particle beam, a pattern is projected on the sample surface from a first direction. In step S30, using the second charged- particle beam, the projected pattern is imaged, in more general words monitored or observed, from a second direction. In step S40, a tilt angle of the sample surface is determined based on the pattern and the imaged pattern. The method ends in step S70.

[0209] Fig. 12 schematically illustrates a triangulation technique for determining the angle a, that can be applied in the dual-beam devices of Figs. 1-7, according to various examples.

[0210] Using the method of pattern projection and marking distance measurement, the FIB-SEM angle a can be precisely measured. This is particularly necessary for variable FIB-SEM arrangements where the angle between FIB and SEM can be changed and is not initially known precisely.

[0211] The measurement is performed as follows: A large field of view of the SEM is set (for example, 200 pm) and a first point Pl, which corresponds to a first marking as disclosed herein, is written on the sample surface using the FIB, ideally in the center of the SEM image. A first SEM image is then taken. The wafer stage is then moved in z-direction (for example, by Az = 50 pm), and a second point P2, which corresponds to a second marking as described herein, is written. A second SEM image is then taken.

[0212] With the change in the x image coordinate of Pl in the first SEM image and P2 in the second image Ax, the FIB-SEM angle a can now be calculated using the following formula:

[0213] 4% a = arctan (— )

[0214] Zlz; By using the image coordinate Pl from the first SEM image (instead of Pl' from the second image), any tilt of the stage z-axis relative to the optical axis of the SEM may be eliminated from the measurement.

[0215] To avoid an unknown tilt of the sample surface being considered as an error in the determination of angle a, the procedure is repeated after a compucentric rotation of the stage by 180° around point Pl has been performed, to arrive at a first angle al before and a second angle a2 after rotation.

[0216] The desired value a is obtained from the average of the two measurements of al and a2. Ideally, this measurement is performed at the wafer center, so that when the stage is rotated, it automatically rotates compucentrically.

[0217] To increase the accuracy of the measurement, this measurement can be repeated at different z positions of the stage, and an average can be calculated over all measurements. This way, linearity errors of the stage are eliminated from the determination of angle a through averaging.

[0218] FIG. 13 is a flowchart of a method for determining an angle a between a first and a second charged-particle beams of a dual-beam device, according to various examples.

[0219] The method starts in step T10. In step T20, using the first charged-particle beam, a first marking is applied at a first marking position on a sample surface, wherein the first charged-particle beam is in a first beam position and the sample is in a first sample position. In step T30, the second charged-particle beam is scanned across the sample surface to acquire a first image of the sample surface including the first marking while the sample is in the first sample position. In step T40, the sample is moved in the second direction by a distance Az to a second sample position. In step T50, using the first charged-particle beam, a second marking is applied at a second marking position on the sample surface, wherein the first charged-particle beam is in the first beam position and the sample is in the second sample position. In step T60, the second charged-particle beam is scanned across the sample surface to acquire a second image of the sample surface including the second marking (imaging first marking may be optional) while the sample is in the second sample position. In step T70, a first imaged marking position of the first marking in the first image and a second imaged marking position of the second marking in the second image are determined. In step T80, the angle a is determined based on the first and second imaged marking positions. The method ends in step T90.

[0220] Although the disclosure has been shown and described with respect to certain preferred embodiments, equivalents and modifications will occur to others skilled in the art upon the reading and understanding of the disclosure. The present disclosure includes all such equivalents and modifications and is limited only by the scope of the appended claims.

[0221] For illustration, above, scenarios have been disclosed in which the wafer tilt is determined with respect to a machine coordinate system of the DBD. In general, using the disclosed techniques, various representations of the wafer tilt, e.g., the wafer tilt with respect to an imaging plane or with respect to a stage tilt may be determined. The respective transformations are readily available.

[0222] Implementations of the subject matter and the functional operations described in this disclosure can be carried out by a dual-beam device including a processing device, for example they can be implemented in digital electronic circuitry, in tangibly-embodied computer software or firmware, in computer hardware, including the structures disclosed in this specification and their structural equivalents, or in combinations of one or more of them. Implementations of the subject matter described in this specification can be implemented as one or more computer programs, i.e., one or more modules of computer program instructions encoded on a tangible program carrier for execution by, or to control the operation of, a processing device. Alternatively or in addition, the program instructions can be encoded on a propagated signal that is an artificially generated signal, e.g., a machine-generated electrical, optical, or electromagnetic signal that is generated to encode information for transmission to suitable receiver apparatus for execution by a processing device. A machine-readable medium can be a machine-readable storage device, a machine- readable storage substrate, a random or serial access memory device, or a combination of one or more of them. The term “processing device” encompasses all kinds of apparatus, devices, and machines for processing information, including by way of example a programmable processor, a computer, or multiple processors or computers. The apparatus can include special purpose logic circuitry, e.g., an FPGA (field programmable gate array) or an ASIC (application-specific integrated circuit) or RISC (reduced instruction set circuit). The apparatus can also include, in addition to hardware, code that creates an execution environment for the computer program in question, e.g., code that constitutes processor firmware, a protocol stack, an information base management system, an operating system, or a combination of one or more of them.

[0223] A computer program (which may also be referred to as a program, software, a software application, a script, or code) can be written in any form of programming language, including compiled or interpreted languages, or declarative or procedural languages, and it can be deployed in any form, including as a stand-alone program or as a module, component, subroutine, or other unit suitable for use in a computing environment. A computer program may, but need not, correspond to a file in a file system. A program can be stored in a portion of a file that holds other programs or information (e.g., one or more scripts stored in a markup language document), in a single file dedicated to the program in question, or in multiple coordinated files (e.g., files that store one or more modules, sub-programs, or portions of code). A computer program can be deployed to be executed on one computer or on multiple computers that are located at one site or distributed across multiple sites and interconnected by a communication network.

[0224] The processes and logic flows described in this specification can be performed by one or more programmable computers executing one or more computer programs to perform functions by operating on input information and generating output. The processes and logic flows can also be performed by, and apparatus can also be implemented as, special purpose logic circuitry, e.g., an FPGA (field programmable gate array) or an ASIC (application-specific integrated circuit) or RISC.

[0225] Computers suitable for the execution of a computer program include, by way of example, general or special purpose microprocessors or both, or any other kind of central processing unit.

[0226] Generally, a central processing unit will receive instructions and information from a read-only memory or a random-access memory or both. The essential elements of a computer are a central processing unit for performing or executing instructions and one or more memory devices for storing instructions and information. Generally, a computer will also include, or be operatively coupled to receive information from or transfer information to, or both, one or more mass storage devices for storing information, e.g., magnetic, magneto-optical disks, or optical disks. However, a computer need not have such devices.

[0227] Computer-readable media (e.g., one or more machine readable hardware storage devices) suitable for storing computer program instructions and information include all forms of non-volatile memory, media and memory devices, including by way of example semiconductor memory devices, e.g., EPROM, EEPROM, and flash memory devices; magnetic disks, e.g., internal hard disks or removable disks; magneto-optical disks; and CD-ROM and (Blue Ray) DVD-ROM disks. The processor and the memory can be supplemented by, or incorporated in, special purpose logic circuitry.

[0228] Implementations of the subject matter described in this specification can be implemented in a computing system that includes a back-end component, e.g., as an information server, or that includes a middleware component, e.g., an application server, or that includes a front-end component, e.g., a client computer having a graphical user interface or a Web browser through which a user can interact with an implementation of the subject matter described in this specification, or any combination of one or more such back-end, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital information communication, e.g., a communication network. Examples of communication networks include a local area network (“LAN”) and a wide area network (“WAN”), e.g., the Internet.

[0229] The computing system can include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. In another example, the server can be in the cloud via cloud computing services. While this specification contains many specific implementation details, these should not be construed as limitations on the scope of any of what may be claimed, but rather as descriptions of features that may be specific to particular implementations. Certain features that are described in this specification in the context of separate implementations can also be implemented in combination in a single implementation. Conversely, various features that are described in the context of a single implementation can also be implemented in multiple implementations separately or in any suitable subcombination. Moreover, although features may be described above as acting in certain combinations and even initially claimed as such, one or more features from a claimed combination can in some cases be excised from the combination, and the claimed combination may be directed to a subcombination or variation of a subcombination.

[0230] Similarly, while operations are depicted in the drawings in a particular order, this should not be understood as requiring that such operations be performed in the particular order shown or in sequential order, or that all illustrated operations be performed, to achieve desirable results. In certain circumstances, multitasking and parallel processing may be used. Moreover, the separation of various system components in the implementations described above should not be understood as requiring such separation in all implementations, and it should be understood that the described program components and systems can generally be integrated together in a single software product or packaged into multiple software products.

[0231] Particular implementations of the subject matter have been described. Other implementations are within the scope of the following claims. For example, the actions recited in the claims can be performed in a different order and still achieve desirable results. As one example, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In certain implementations, multitasking and parallel processing may be used.

[0232] From the above said, the following general conclusions can be drawn:

[0233] In general, the method may comprise the step of projecting a pattern onto a sample surface by a first charged-particle beam from a first direction, and the step of imaging the projected pattern by a second charged-particle beam from a second direction, and the step of determining a tilt angle based on the projected pattern and the imaged pattern.

[0234] In such a pattern, it can be sufficient to write a pair of points for each direction (x and y) and measure the resulting distance between the points in the SEM image. However, a large number of points can also be written, and the average value of the point distances measured in the SEM image (in both x and y) can be used to determine the mean tilt angle. This makes the determination of the tilt angle more reliable.

[0235] With a sufficient number of individual points, the topography of the wafer surface can be measured, and local curvatures in the imaged wafer region can also be determined. The points written with the FIB should be as small as possible to minimize changes to the sample. To still detect them reliably in the SEM image, the written point can be detected in an SEM difference image before and after writing the point, and the point center can be determined, for example, by center of gravity determination or by fitting a circle to the point edges.

[0236] For example, in an SEM image with a pixel resolution of 1 nm and an FIB-SEM angle of 25°, a change in point distance of 25 pixels in the SEM image per milliradian (mrad) of tilt angle Q results for an FIB beam distance of 5 pm.

[0237] In various examples, by writing further points, in each tilt direction, generally referred to as markings, the mean tilt angle can be determined more reliably by averaging the point distances in the SEM image. By writing a large number of points on the wafer surface using the FIB, the tilt angle may be determined more reliably. Instead of relying on a single pair of points for each direction (x and y), multiple points may be written, and the average value of the measured point distances in the SEM image may be used to calculate the mean tilt angle. This approach may reduce the impact of any individual measurement errors and may provide a more statistically robust estimation of the tilt angle.

[0238] In various examples, applying a plurality of further markings onto the sample surface positioned in a 2D pattern using the first charged-particle beam may be included. A plurality of further tilt angles may be determined based on the further markings. The first and / or second tilt angle may be determined further based on a statistical analysis of the further tilt angles. By using a larger number of markings arranged in a two-dimensional pattern, the tilt angles can be determined with higher accuracy and precision, as the statistical analysis of the multiple tilt angle measurements can reduce the impact of individual measurement errors.

[0239] The point pattern written by the FIB may be either equidistant or non-equidistant. An equidistant pattern means that the spacing between the points may be uniform in both x and y directions. This may simplify the calculations and may provide a consistent sampling of the wafer surface. However, a non-equidistant pattern, where the spacing between points varies, may also be used. This flexibility may allow for adapting the point pattern to specific regions of interest or accommodating any constraints in the FIB writing process.

[0240] The spacings between the further markings in the 2D pattern may be equidistant in both directions of the 2D pattern on the sample surface. Alternatively, the spacings between the further markings in the 2D pattern may vary along at least one direction on the sample surface. The choice of equal or varying spacings may depend on the specific sample characteristics and the desired spatial resolution of the tilt angle determination.

[0241] The markings may be point markings, cross markings, or other geometric markings that allow precise determination of the marking center, for example rotationally or axially symmetric markings. For example, the point center may be determined using methods like center of gravity calculation or fitting a circle to the point edges.

[0242] To minimize the impact on the sample surface, it may be advantageous to write the smallest possible markings with the FIB, e.g., in the size of the semiconductor structures to be imaged.

[0243] A SEM difference image technique may be employed. An SEM image of the wafer surface may be captured before and after writing each marking. By subtracting these two images, a difference image may be obtained, which may highlight the changes caused by the written point. This may enhance the contrast and may make it easier to detect and locate the point accurately. Before applying the first marking and / or the second marking, the method may include imaging the sample surface to acquire a reference image, one reference image for both markings, or for each marking a separate respective reference image, of sample surface. A difference image may be generated based on the reference image and the image, wherein the difference image represents modifications of the sample surface due to the applied first and / or second markings. The first and / or second imaged marking positions of the first and second markings may be determined using the difference image. By comparing the reference image (before marking) with an image after marking, for each marking individually, the positions of the markings can be more precisely identified.

[0244] The method may include depositing a protective layer on the sample surface using a gas injection system (GIS) before applying the first and second markings. This layer may be deposited locally or over a larger area using a gas injection system (GIS) in combination with the SEM or FIB. The protective layer may provide a flat and thin surface on which the point pattern may be written, preserving the integrity of the underlying wafer surface.

[0245] The first and second markings may then be applied onto the protective layer. The protective layer serves to prevent damage to the sample surface during the marking process and may also enhance the visibility of the markings in the acquired images.

[0246] In some cases, the tilt angle measurement may not be performed directly within the ROI intended for the 3D tomography process. Assuming that the mean tilt angle varies slowly across the wafer surface, the tilt measurement may be carried out a few micrometers away from the ROI. This approach may allow for tilt angle determination without interfering with the specific area of interest, while still providing a representative measurement of the local tilt. Therefore, the first, second, and optionally third markings may be applied onto the sample surface in a marking region adjacent to a region of interest (ROI) to be imaged. By placing the markings close to, but outside of, the ROI, the tilt angles can be determined without interfering with the imaging of the ROI itself.

[0247] While the described method primarily focuses on writing point patterns, it may not be limited to that. Line patterns may also be written using the FIB, enabling the measurement of the wafer topography along the written lines. By writing multiple line patterns and applying triangulation techniques, a detailed topographical map of the wafer surface may be obtained. This extension of the method may provide valuable information about the surface contours and any local variations in height.

[0248] The two-dimensional pattern may, for example, be a regular array, such as a rectangular grid or a hexagonal grid, or an irregular array with varying spacings between the markings. The method may further comprise determining a curvature of the sample surface based on the further tilt angles. By analyzing the variation of the tilt angles across the 2D pattern, the local curvature of the sample surface can be estimated, providing additional information about the sample topography.

[0249] The tilt angle measurement method may be further utilized to determine the exact 3D contour of the wafer region processed by 3D tomography. By writing point or line patterns strategically within the processed volume, the method may accurately measure the surface topography at different stages of the tomography process. This may enable the determination of the total cutting depth achieved during the 3D tomography, providing valuable information for process control and analysis.

[0250] The method may include performing a 3D tomography of the sample and reconstructing a 3D volume data set of the sample using the determined first and / or second tilt angle to correct for perspective distortion. By incorporating the tilt angle information into the 3D reconstruction process, the spatial accuracy of the resulting 3D volume data set can be improved, providing a more faithful representation of the sample's three-dimensional structure.

[0251] These advantageous embodiments highlight the versatility and robustness of the described tilt angle measurement method. By incorporating techniques like writing multiple points, using difference images, depositing protective layers, measuring near the ROI, writing line patterns, and extending the method to 3D contour and FIB-SEM angle measurements, the overall accuracy, reliability, and applicability of the method can be significantly enhanced.

[0252] The method may further include determining the first tilt angle by applying an inverse trigonometric function. Such as arccosine or arcsine, for example to the ratio of the known beam distance to the measured distance between markings. This approach directly calculates the tilt angle using the geometric relationship between the beam distance and the observed marking separation.

[0253] In some embodiments, the trigonometric relationship used to calculate the tilt angle may account for small angle approximations when the sample tilt is determined to be below a predetermined threshold angle. These approximations simplify the calculations by replacing trigonometric functions with their linear approximations, which are valid for small angles.

[0254] Determining the second tilt angle, which represents the tilt around an axis perpendicular to the first tilt axis, may involve calculating a tangent-based relationship between the first tilt angles obtained from two separate pairs of markings. This relationship utilizes the difference in the first tilt angles and the known beam distances to estimate the tilt in the perpendicular direction.

[0255] The method may further comprise iteratively refining the tilt angle calculation by applying progressively more precise trigonometric approximations. This iterative approach starts with simple approximations and gradually incorporates higher-order terms to improve the accuracy of the tilt angle determination, particularly for larger tilt angles.

[0256] The determination of tilt angles may involve solving a system of trigonometric equations that relate the known beam geometry, including the beam distances and the angular relationship between the charged-particle beams, to the observed positions of the markings in the image captured by the second charged-particle beam. This approach provides a comprehensive solution for the tilt angles by simultaneously considering all available geometric information. In various examples, other approaches may be used to determine the tilt angle of the sample surface based on the projected and imaged patterns, in addition to the trigonometric approach. The projected and imaged patterns may be analyzed to extract specific features, like edges, corners, or centroid positions of pattern elements. The spatial relationships (e.g., distances, angles) between these features may computed for both patterns. The differences in these spatial relationships are then be used to calculate the tilt angle using geometric transformations. In another approach, a set of possible tilt angles may be hypothesized, and for each angle, the expected appearance of the imaged pattern is simulated by applying a geometric transformation (e.g., affine or perspective) to the projected pattern. The simulated imaged patterns are then compared with the actual imaged pattern using a similarity metric (e.g., correlation, mutual information). The tilt angle that yields the highest similarity between the simulated and actual imaged patterns is selected as the estimated tilt angle.

[0257] In another approach, a machine learning model, such as a convolutional neural network (CNN), may trained on a dataset of projected patterns and their corresponding imaged patterns at known tilt angles. The trained model may then be used to predict the tilt angle directly from new pairs of projected and imaged patterns, based on the learned mapping between pattern distortions and tilt angles.

[0258] These alternative approaches offer different ways to process and analyze the projected and imaged patterns, while other approaches are possible.

[0259] Summarizing, a method for determining sample tilt in a dual-beam metrology device is provided. The dual-beam device comprises a first charged-particle beam for processing the sample and a second charged-particle beam for imaging the sample, which are arranged at an angle a relative to each other with regard to their direction to the sample. The method comprises applying first and second markings on the sample surface using the first charged-particle beam, with a known beam distance between the marking positions in perpendicular direction to the first charged- particle beam. The second charged-particle beam acquires an image of the markings. The marking positions in the acquired image are determined, and a tilt angle of the sample relative to the second beam can be calculated based on the angle a, the beam distance, and the imaged marking positions. The disclosed techniques may be used to calibrate the angle a in a variable dual-beam device. The disclosed techniques enable real-time tilt correction of images based on the imaging and milling processing modules, thereby improving accuracy of 3D tomography reconstruction.

Claims

CLAIMS1. A method for determining tilt of a sample in a dual-beam metrology device, wherein the dual-beam metrology device comprises a first charged-particle beam arranged in a first direction towards the sample, and a second charged-particle beam arranged in a second direction towards the sample, wherein the first and second directions define an angle a relative to each other, comprising the following steps:- projecting, from the first direction using the first charged-particle beam, a pattern onto a sample surface;- imaging, from the second direction using the second charged-particle beam, the projected pattern on the sample surface; and- determining a tilt angle of the sample surface based on the projected pattern and the imaged pattern.

2. The method according to claim 1, wherein projecting a pattern onto the sample surface comprises:- applying, using the first charged-particle beam, a first marking at a first marking position on the sample surface, wherein the first charged-particle beam is in a first beam position; and- applying, using the first charged-particle beam, a second marking at a second marking position on the sample surface, wherein the first charged-particle beam is in a second beam position, wherein the first charged-particle beam in the first beam position is distanced from the charged-particle beam in the second beam position by a first beam distance dx at the sample; wherein imaging the projected pattern comprises:- scanning the second charged-particle beam across the sample surface to acquire an image of the sample surface including the first and second markings; and wherein determining a tilt angle comprises:- determining a first imaged marking position of the first marking in the image and a second imaged marking position of the second marking in the image; and- determining a first tilt angle of the sample surface relative to the imaging beam along a first tilt direction between the first and second markings based on the angle a, the firstbeam distance dx, and the first and second imaged marking positions.

3. The method according to claim 2, wherein determining the sample first tilt angle comprises computing at least one trigonometric formula based on the angle a.

4. The method according to one of the preceding claims, wherein the first charged- particle beam originates from a Focused Ion Beam (FIB) module, and the second charged-particle beam originates from a Scanning Electron Microscope (SEM) module, of the dual-beam metrology device.

5. The method according to one of claims 2 to 4, wherein determining the first and second imaged marking positions in the image comprises determining a center position of respective marking in the image.

6. The method according to one of claims 2 to 5, further comprising:- before applying the first marking and / or the second marking, imaging the sample surface, in order to acquire a reference image of the sample surface;- generating a difference image based on the reference image and the image, wherein the difference image represents modifications of the sample surface due to the applied first and / or second markings; and- determining the first and / or second imaged marking positions of the first and second markings using the difference image.

7. The method according to one of claims 2 to 6, further comprising:- applying, using the first charged-particle beam, a third marking at a third marking position on the sample surface, wherein the first charged-particle beam is in a third beam position; wherein the first charged-particle beam in the third beam position is distanced from the first charged-particle beam in the first beam position by a second beam distance dy in a perpendicular direction relative to the first beam distance dx;- applying, using the first charged-particle beam, a fourth marking at a fourth marking position on the sample surface, wherein the first charged-particle beam is in a fourth beam position;wherein the first charged-particle beam in the fourth beam position is distanced from the charged-particle beam in the third beam position by the first beam distance dx at the sample;- determining a third and a fourth imaged marking position of the third respectively fourth marking in the image; and- determining a second tilt angle of the sample surface in a second tilt direction perpendicular to the first tilt direction between the first and third markings further based on the second beam distance dy, the third and fourth imaged marking positions.

8. The method according to one of claims 2 to 7, further comprising:- depositing a protective layer on the sample surface using a gas injection system (GIS) before applying the markings to the sample surface, wherein the markings are applied onto the protective layer.

9. A method for determining an angle a between a first and a second charged-particle beams of a dual-beam device, wherein the first charged-particle beam is arranged in a first direction towards a sample, and the second charged-particle beam is arranged in a second direction towards the sample, wherein the first and second directions define the angle a relative to each other, the method comprising:- applying, using the first charged-particle beam, a first marking at a first marking position on a sample surface, wherein the first charged-particle beam is in a first beam position and the sample is in a first sample position;- scanning the second charged-particle beam across the sample surface to acquire a first image of the sample surface including the first marking while the sample is in the first sample position;- moving the sample in the second direction by a distance Az to a second sample position;- applying, using the first charged-particle beam, a second marking at a second marking position on the sample surface, wherein the first charged-particle beam is in the first beam position and the sample is in the second sample position;- scanning the second charged-particle beam across the sample surface to acquire a second image of the sample surface including the first and second markings while the sample is in the second sample position;- determining a first imaged marking position of the first marking in the first image and asecond imaged marking position of the second marking in the second image;- determining the angle a at least based on the first and second imaged marking positions.

10. A dual-beam metrology device, comprising: - a first device configured to generate a first charged-particle beam arranged in a first direction towards a sample;- a second device configured to generate a second charged-particle beam arranged in a second direction towards the sample, wherein the first and second directions define an angle a relative to each other; and - a control device configured to control the first device and the second device and to carry out the method of one of claims 1 to 9.

Images