Charged particle beam apparatus and three-dimensional shape measurement method

Abstract

This invention is robust with respect to measurement conditions or target samples, and enables three-dimensional shape measurement with high measurement accuracy. This charged particle beam device includes: a detector group comprising a plurality of detectors 105 arranged in mutually different directions; and a processor. The processor generates a calibration parameter corresponding to WD when a sample is imaged by the detector group, and generates a three-dimensional image of the sample on the basis of the generated calibration parameter. With respect to the output from the detector group, the processor calculates a three-dimensional reconstruction in which a condition corresponding to a height map of the sample is applied for each region in the observation field of view, and generates a three-dimensional image of the sample.

Description

[Technical Field]

[0001] The present invention relates to a charged particle beam apparatus and a three-dimensional shape measurement method using the same. [Background Art]

[0002] Patent Document 1 discloses a scanning electron microscope that can easily perform three-dimensional shape measurement. The scanning electron microscope comprises: a plurality of divided backscattered electron detectors each detecting backscattered electrons from a sample surface; a signal switching unit that outputs detection signals for the number of divisions from the backscattered electron detectors while switching the detection signals for each scanning unit; and a memory that records the detection signals output from the signal switching unit and has storage areas for the number of divisions. Using this configuration, the scanning electron microscope acquires image data of scanning units for the number of divisions, and performs arithmetic processing for three-dimensional shape measurement based on the positions of the backscattered electron detectors and the intensity of the detection signals.

[0003] Patent Document 2 discloses a method for calculating a three-dimensional topographical model of a sample surface from a plurality of SEM images of the sample. In this method, a primary electron beam is irradiated onto the surface of a sample composed of a plurality of material types, and emitted electrons emitted from the sample are detected using a plurality of detectors arranged at different positions relative to the sample. Then, horizontal regions on the surface are identified for each material type, and a calibration coefficient for each material type is calculated based on at least one detector output in the identified horizontal regions. The calibration coefficient compensates for variations in emitted electron yield between material types, and a three-dimensional topographical model of the surface is calculated based on the detector output to which the calibration coefficient is applied. [Prior Art Documents] [Patent Documents]

[0004] [Patent Document 1] Japanese Unexamined Patent Publication No. 2008-282761 [Patent Document 2] Japanese Unexamined Patent Publication No. 2015-8148 [Summary of the Invention] [Problems that the invention aims to solve]

[0005] Charged particle beam microscopes (SEMs) scan the surface of a sample with a finely focused beam of charged particles, detecting the signals emitted from the sample to form a two-dimensional profile of the sample surface, and displaying it as an image. While high-resolution images can be obtained in the scanning direction, information perpendicular to the scanning direction, i.e., in the sample height direction, is usually not obtained. Several methods have been developed to obtain the three-dimensional structure of a sample using charged particle beam microscopes.

[0006] For example, one method involves thinning the sample into multiple sections using an ultramicrotome, observing each section with a SEM to obtain stacked images, and then performing three-dimensional reconstruction. Another method involves using a FIB-SEM to thinly shave a target area of ​​the sample with a focused ion beam, observing it with a SEM, and repeating this process to obtain stacked images and perform three-dimensional reconstruction. While these methods offer high resolution in the height direction and allow for analysis of the internal structure of the sample, they also destroy the sample, require large-scale equipment, and demand enormous processing time.

[0007] Non-destructive methods for obtaining 3D morphological information of a sample surface include the stereo method, which involves tilting the sample or changing the angle of incidence of a particle beam to acquire multiple images with parallax angles, and then calculating height information from the correspondence between the feature points of each image. Other methods have been developed, such as acquiring multiple images with different focal positions to determine the focal position and create a depth map, or calculating the normal to the sample surface from shaded images obtained by a detector array consisting of multiple detectors arranged in different directions to determine the sample height.

[0008] While these methods for obtaining surface morphology can be performed quickly using relatively simple configurations, they suffer from poor height accuracy and numerous artifacts. Among these methods, the last one mentioned, which uses multiple detectors arranged in different directions, has been widely used, and improvements are being made, such as its application to general-purpose scanning electron microscopes as shown in Patent Document 1 and its application to semiconductor device measurement as shown in Patent Document 2.

[0009] However, while Patent Document 1 enables three-dimensional shape measurement with a simple configuration that suppresses errors due to detection signal saturation, it does not address further improvements in accuracy. Patent Document 2 primarily focuses on semiconductor devices and is not able to accommodate a wide range of observation targets or the wide range of observation conditions required for them.

[0010] In the method of determining sample height by calculating the normal to the sample surface from shaded images obtained by multiple detectors positioned in different directions, it is utilized that the shading caused by the inclination of the sample surface differs depending on the positional relationship between the sample and the detector. Therefore, for example, if the working distance (abbreviated as WD) deviates from the specified position, or if the unevenness of the sample is large and the positional relationship between the sample surface and the detector changes depending on the sample position, it will affect the shaded image, and thus the height calculation, becoming a source of error in three-dimensional shape measurement.

[0011] Such errors are negligibly small in applications such as length-measuring SEMs where the working width (WD) is nearly fixed, and for samples like semiconductor devices where the surface is considered nearly flat due to pattern-level irregularities, or under high-magnification conditions where the field of view is narrow and the height difference within the observation area is limited. However, such errors significantly affect measurement accuracy in instruments used for a wide range of applications, including particle analysis, surface topography analysis, and fracture surface analysis.

[0012] Therefore, one of the objectives of the present invention is to enable three-dimensional shape measurement with high measurement accuracy that is robust to measurement conditions or target samples. [Means for solving the problem]

[0013] To solve the above problems, one embodiment provides a charged particle beam apparatus comprising a detector group consisting of a plurality of detectors arranged in different directions from one another, and a processor, wherein the processor generates calibration parameters corresponding to the WD when the sample is imaged by the detector group, and generates a three-dimensional image of the sample based on the generated calibration parameters.

[0014] In one embodiment, a charged particle beam apparatus is provided which includes a detector group consisting of multiple detectors arranged in different directions from one another, and a processor, wherein the processor performs a three-dimensional reconstruction calculation on the output from the detector group, applying conditions according to the height map of the sample for each region within the observation field, thereby generating a three-dimensional image of the sample. [Effects of the Invention]

[0015] According to one embodiment, it is possible to perform three-dimensional shape measurement with high measurement accuracy, which is robust to measurement conditions or target samples.

[0016] Further features related to the present invention will become apparent from the description herein and the accompanying drawings. Problems, configurations, and effects other than those described above will be clarified by the following description of the embodiments. [Brief explanation of the drawing]

[0017] [Figure 1] This is a schematic diagram showing an example of the overall configuration of a charged particle beam apparatus according to the first embodiment. [Figure 2] Figure 1 is a schematic diagram illustrating an example of the signal detection principle when performing three-dimensional shape measurement using a scanning electron microscope. [Figure 3] Figure 1 is a schematic diagram showing an example of the detector arrangement and an example of a shadow image acquired by the detector. [Figure 4] Figure 3 shows an image from a detector and an example of a three-dimensional shape display based on that image. [Figure 5] This figure shows an example of a three-dimensional shape profile based on the height map shown in Figure 4. [Figure 6] This figure illustrates the difference in signal detection due to differences in working distance (WD) when performing three-dimensional shape measurement using the scanning electron microscope shown in Figure 1 in the first embodiment. [Figure 7] This flowchart shows an example of the processing content of a charged particle beam apparatus according to the first embodiment. [Figure 8]It is a diagram showing an example of detection characteristics of the detector in FIG. 3 and FIG. 6. [Figure 9] In the second embodiment, it is a diagram for explaining a difference in signal detection due to a difference in samples when three-dimensional shape measurement is performed using the scanning electron microscope shown in FIG. 1. [Figure 10] It is a flowchart showing an example of processing content of the charged particle beam apparatus according to the second embodiment. [Figure 11] It is a schematic diagram showing an example of the overall configuration of the charged particle beam apparatus according to the third embodiment. [Figure 12] It is a flowchart showing an example of processing content of the charged particle beam apparatus according to the third embodiment. [Figure 13] In the fourth embodiment, it is a schematic diagram showing an example of the shape of a sample to be measured. [Figure 14] In the charged particle beam apparatus according to the fourth embodiment, it is a flowchart showing an example of processing when performing three-dimensional reconstruction. MODE FOR CARRYING OUT THE INVENTION

[0018] Hereinafter, the present embodiment will be described with reference to the accompanying drawings. In the accompanying drawings, functionally identical elements may be indicated by the same number or corresponding numbers.

[0019] (First Embodiment) <Outline of Charged Particle Beam Apparatus> FIG. 1 is a schematic diagram showing an example of the overall configuration of the charged particle beam apparatus according to the first embodiment. Here, a case where charged particles are electrons, that is, a case where the charged particle beam apparatus is a scanning electron microscope (SEM) will be described as an example. As an example, the scanning electron microscope shown in FIG. 1 is generally configured of a sample chamber 101, an electron optical system 102, a sample stage 103, a stage 104, a detector 105, an imaging device 110, an arithmetic control unit 120, and a stage control device 130.

[0020] The sample chamber 101 has a function to maintain a vacuum inside so that the electron beam does not scatter, and is configured to accommodate a sample inside. The electron optical system 102 consists of an electron source 161 that generates an electron beam such as electrons, a condenser lens 162 that narrows the electron beam emitted from the electron source 161, a deflector 163 that deflects the electron beam, and an objective lens 164 that focuses the electron beam.

[0021] The optical system control device 170 includes a memory 171 for storing various parameters related to the control of the electron optical system 102. Based on these parameters, the optical system control device 170 generates an electron beam from the electron source 161. The optical system control device 170 also has the function of controlling the electron beam from the electron source 161 to be focused by a condenser lens 162, deflected and focused by a deflector 163 and an objective lens 164, and irradiated onto a sample 108 mounted on a sample stage 103, based on these parameters. The configuration in Figure 1 is an example, and the electron optical system 102 may include other lenses and electrodes in addition to the elements shown, or may be replaced with similar elements, and the details of the configuration are not limited to those shown.

[0022] The sample stage 103 is mounted on the stage 104. The stage 104 is movable in the XYZ directions (vertical, horizontal, and height directions) within the sample chamber 101. The stage 104 also holds the sample stage 103 so that it can rotate around a rotation axis and tilt with the X or Y direction as the tilt axis. The stage 104 is controlled by the stage control device 130 according to processing in the calculation control unit 120. The stage control device 130 includes a coordinate storage unit 131 that stores the coordinates of the stage 104, a control unit 132 that controls the coordinates of the stage 104 according to instructions from the calculation control unit 120, and a drive unit 133 that drives the stage 104 based on signals from the control unit 132.

[0023] Detector 105 has the function of detecting backscattered electrons, secondary electrons, energy-dispersive X-rays (EDS), electron backscatter diffraction (EBSD), wavelength-dispersive X-ray spectroscopy (WDS), cathodoluminescence (CL), Auger electrons, and electron energy loss spectroscopy (EELS) emitted from sample 108. Although only one detector 105 is shown in Figure 1, it goes without saying that the number and position of detectors 105 are not limited to a specific configuration. The imaging device 110 has the function of converting the signals detected by detector 105 into images and has an internal memory 111 for storing signal information.

[0024] The arithmetic control unit 120 controls various components of the scanning electron microscope and controls the display of observation results according to observation conditions and other inputs from the user. The arithmetic control unit 120 is composed of an information processing device such as a computer. For example, the arithmetic control unit 120 includes a CPU (processor) 121, a main memory device 122 such as memory, and a secondary memory device 123 such as a hard disk drive (HDD) or solid state drive (SSD). The arithmetic control unit 120 also includes an input unit 124 such as a keyboard, mouse, or touch monitor, a display unit 125 such as a liquid crystal display, and a communication unit (not shown) that communicates with the various components of the scanning electron microscope. The display unit 125 includes an observation image display unit 1251 and a 3D image display unit 1252.

[0025] The main memory 122 stores a computer program that controls the operation of the entire scanning electron microscope. The computer program provides functional blocks such as the information processing unit 1221 and the display processing unit 1224 within the main memory 122. The information processing unit 1221 is a processing unit that processes information acquired by the optical system control device 170, such as acceleration voltage, emission current, magnification, beam spot and working distance (WD), information acquired by the imaging device 110, such as signal, image name, image resolution, field of view, and imaging time, and information acquired by the stage control device 130, such as stage coordinates. The information processing unit 1221 outputs the processed information as image attribute information. Furthermore, the information processing unit 1221 performs processing to calculate the three-dimensional shape of the sample 108 from the image data and attribute information acquired by the imaging device 110.

[0026] The display processing unit 1224 performs display processing on the acquired image of the sample 108 based on attribute information at the time of imaging, such as stage coordinates, stage rotation angle, magnification, and raster rotation, and displays the processed image on the observation image display unit 1251. Furthermore, the display processing unit 1224 performs display processing on the three-dimensional shape measurement results and displays the processed image on the 3D image display unit 1252.

[0027] <Basic principles of three-dimensional shape measurement> Figure 2 is a schematic diagram illustrating an example of the signal detection principle when performing three-dimensional shape measurement using the scanning electron microscope shown in Figure 1. Figure 3 is a schematic diagram illustrating an example of the detector arrangement configuration in Figure 1 and an example of a shadow image acquired by the detector. In Figure 2, when an electron beam is shone onto a sample 108 on the sample stage 103 (treated here as a thin, flat sample and not shown), backscattered electrons 106 are emitted. These backscattered electrons 106 are detected by a detector 105 positioned directly below the objective lens 164, converted into an electrical signal, and then an image based on this electrical signal is formed.

[0028] Figure 3 shows an example of the detector 105 as viewed from the sample surface (below). More specifically, the detector 105 is composed of a detector group consisting of multiple detectors that are divided and arranged in different directions from one another. In this example, the detector group consists of four detectors 105a to 105d arranged in a ring with respect to the electron beam path. Backscattered electrons are generated when the electron beam that has passed through the central opening is irradiated onto the sample. The four detectors 105a to 105d, which are arranged in a ring, or in other words, symmetrically with respect to the irradiation axis of the electron beam, each detect the component of the backscattered electrons that reaches the ring region. That is, the four detectors 105a to 105d detect the backscattered electrons separately for each emission angle.

[0029] Backscattered electrons have a cosine emission intensity with respect to the azimuth angle of the sample. This is shown in Figure 2 by the arrows within the dashed line. Although backscattered electrons 106 are emitted in all directions, Figure 2 shows only the component detected by detector 105. In observation state [1] in Figure 2, the sample is placed flat with respect to the electron beam, so backscattered electrons 106 are detected uniformly by detectors 105a to 105d. On the other hand, in observation state [2] in Figure 2, the sample is placed at an angle with respect to the electron beam, so the detection signal amounts of detectors 105a to 105d differ depending on the direction of the sample's tilt.

[0030] Figure 3 schematically shows examples of images IMGa to IMGd obtained by observing Vickers marks, which are indentations created by driving a square pyramidal bit into a material, using each detector 105a to 105d. The image density changes depending on the angle of the sample surface; when the surface is highly inclined relative to the detector, the amount of detected signal decreases, resulting in a darker, shaded image. By utilizing this characteristic, the sample surface angle can be estimated by calculating the detected signals from detectors 105a to 105d, converting this to height, and then performing area integral to obtain the three-dimensional shape of the sample surface.

[0031] Figure 4 shows an example of the image obtained by the detector and the three-dimensional shape display based on that image, as shown in Figure 3. In Figure 4, image IMGa is the image obtained when a wire with a circular cross-section is observed by detector 105a. Image IMGc is the image obtained when the same wire is observed by detector 105c, which is located opposite detector 105a. In image IMGa, since detector 105a is positioned to the left of the wire, the right side of the wire is dark and in shadow. Conversely, in image IMGc, the left side of the wire is dark and in shadow.

[0032] The composite image CIMG is an image created by combining image IMGa and image IMGc. In the composite image CIMG, the directionality due to detectors 105a and 105c is eliminated, and no difference in shading between the left and right sides of the wire is observed. Here, the difference ratio Dac is calculated for each pixel by dividing the difference between image IMGa and image IMGc by the sum of image IMGa and image IMGc. The sample plane angle is determined from the difference ratio Dac, and the sample plane angle is converted to pixel height. Then, the height map is obtained by sequentially integrating the converted pixel height in a two-dimensional space consisting of multiple pixels. To determine the sample plane angle from the difference ratio Dac, for example, the difference ratio Dac with respect to the sample plane angle is determined in advance using a specified sample, and the comparison relationship is used.

[0033] Figure 4 shows an example of a three-dimensional display of the height map HMP obtained in this way. The lower part of the wire, i.e., the semicircular part on the sample stage 103 side, which is a blind spot for detectors 105a and 105c, cannot be detected and therefore no information can be obtained, but the upper part of the wire's three-dimensional shape is reproduced well. Figure 5 shows an example of a three-dimensional shape profile based on the height map HMP shown in Figure 4. In Figure 5, the solid line 501 is the cross-sectional profile of the center of the wire, i.e., the cross-sectional profile by the cross-sectional plate shown in the height map HMP in Figure 4. The dashed line 502 simulates a circular wire cross-section.

[0034] <Measures to address working distance (WD)> Figure 6 illustrates the difference in signal detection due to differences in working distance (WD) when performing three-dimensional shape measurement using the scanning electron microscope shown in Figure 1 in the first embodiment. Figure 6 shows the same observation state as in Figure 2 [2]. In Figure 6, the working distance (WD) is an index based on the distance between the sample and the detector group, and more specifically, it is the distance from the bottom surface of the objective lens 164 to the sample surface. d is the distance from the detection surface of the detector 105 to the sample surface, and WDd is the distance from the bottom surface of the objective lens 164 to the detection surface of the detector 105.

[0035] Figure 6 further shows the observation state [3], where the sample is moved upward by the Z mechanism, i.e., the height adjustment mechanism, of stage 104 (figure omitted), with reference to the observation state [2]. As the sample approaches the detection surface, the relative relationship between the detection surface and the sample surface changes. Specifically, as the sample approaches the detection surface, the difference in the amount of detection signal between detectors increases, the shadow becomes stronger, and the difference ratio Dac increases.

[0036] General-purpose scanning electron microscopes are used for a wide variety of applications, and the Z-axis mechanism of the stage allows for a wide range of movement, from a few millimeters to tens of millimeters, and the working width (WD) can also be adjusted over a similarly wide range. The WD can be changed according to the purpose: a short WD with low aberration is used for high-resolution observation; a WD suitable for analytical detectors is used for analysis such as EDS; and a long WD is used for low-magnification observation or observation of samples at a high incline.

[0037] However, as mentioned above, a change in WD causes the difference ratio Dac to change even with the same sample plane angle, resulting in errors in height calculations. Therefore, to accommodate various WDs, it is necessary to determine the relationship between the difference ratio Dac and the sample plane angle (hereinafter referred to as calibration parameters) for each WD and select the appropriate parameter according to the imaging WD. For example, in Figure 5, the cross-sectional profile of solid line 501 was measured at WD=15mm, but if the calibration parameters for WD=10mm are applied to this data, the cross-sectional profile becomes that of dotted line 503. Thus, it can be seen that the cross-sectional shape deviates significantly from the actual shape when the WD changes.

[0038] Therefore, the first embodiment is characterized by deriving calibration parameters according to the WD during imaging and generating a three-dimensional image of the sample based on the derived calibration parameters. Figure 7 is a flowchart showing an example of the processing content of the charged particle beam apparatus according to the first embodiment. In Figure 7, the charged particle beam apparatus starts electron beam irradiation of the sample 108 and performs predetermined image adjustments, etc. (step S71).

[0039] Next, the charged particle beam apparatus detects backscattered electron (BSE) signals using multiple detectors 105 that constitute the detector group, generates images for each of the multiple detectors by image capture by the imaging device 110, and records the imaging conditions, including the imaging WD, as supplementary information to the image (step S72). Subsequently, the charged particle beam apparatus, in particular the arithmetic control unit 120 including a processor, derives calibration parameters corresponding to the imaging WD included in the supplementary information (step S73).

[0040] Finally, the arithmetic control unit 120 performs a three-dimensional reconstruction process using the calibration parameters derived in step S73 and displays the processing result on the display unit 125 (step S74). In other words, the arithmetic control unit 120 generates a three-dimensional image of the sample based on the output from the detector group, i.e., the images for each of the multiple detectors generated in step S72, and the calibration parameters derived in step S73. The arithmetic control unit 120 then displays the generated three-dimensional image of the sample on the 3D image display unit 1252.

[0041] The method for deriving calibration parameters according to the imaging WD in step S73 will be explained in detail. Figure 8 shows an example of the detection characteristics of the detector 105 in Figures 3 and 6. Various calibration methods can be considered depending on the detector configuration and observation condition range, and calibration parameters can also take various forms. As an example, Figure 8 plots the distance d from the detection surface to the sample surface on the horizontal axis and the pixel height on the sample inclined surface where the difference ratio Dac is a specified value (0.3 in this example) in actual measurement on the vertical axis.

[0042] Specifically, for example, the sample inclination surface is changed in advance using a calibration sample, etc., so that the difference ratio Dac between detector 105a and detector 105c becomes a specified value for each distance d, and the pixel height per pixel is determined based on the changed sample inclination surface. In the specification, this pixel height relative to the predetermined difference ratio Dac is called the difference ratio height. Here, in a certain system (device configuration), as shown in Figure 8, the distance d from the detection surface to the sample surface and the difference ratio height are approximately proportional. The d value is the value obtained by subtracting a fixed value (the distance WDd from the bottom surface of the objective lens 164 to the detection surface of the detector 105) from the imaging WD, and is uniquely determined with respect to the imaging WD.

[0043] In this case, by pre-determining the proportionality constant of the proportional relationship shown in Figure 8, it is possible to derive the correction amount to be applied to the difference ratio height, i.e., the calibration parameter, according to the amount of change when the distance d, and consequently the imaging WD, changes. Thus, the calibration parameter is used to determine the correct difference ratio height from the difference ratio Dac according to the imaging WD. From a practical standpoint, for example, even if the imaging WD changes after various adjustments have been made to obtain a correct height map at a certain imaging WD, the height map can be correctly corrected using the calibration parameter.

[0044] Furthermore, as mentioned above, the relationships between these calibration parameters can vary depending on the device configuration. For example, the relationship between imaging WD and difference ratio height may not be proportional to a certain difference ratio Dac, and the relationship between imaging WD and difference ratio height may change depending on the magnitude of the difference ratio Dac. In any case, by determining the relationship between calibration parameters with respect to imaging WD experimentally or by calculation, and storing this relationship as a table or calculation formula in the secondary storage device 123, etc., it is possible to accommodate various device configurations.

[0045] <Main effects of the first embodiment> As described above, in the method of the first embodiment, by deriving calibration parameters according to the imaging working distance (WD), it is possible to perform three-dimensional shape measurement with minimal degradation of accuracy even if the imaging WD changes. In other words, it becomes possible to perform three-dimensional shape measurement that is robust to measurement conditions and has high measurement accuracy.

[0046] (Second Embodiment) <Measures to address sample height> Figure 9 illustrates the difference in signal detection due to differences in samples when performing three-dimensional shape measurement using the scanning electron microscope shown in Figure 1 in the second embodiment. The observation state shown in Figure 9 [4] is the same as the observation state including the flat sample stage 103 shown in Figure 2 [1]. As in the case of Figure 6, the working distance (WD) is the distance from the bottom surface of the objective lens 164 to the sample surface. d is the distance from the detection surface of the detector 105 to the sample surface, and WDd is the distance from the bottom surface of the objective lens 164 to the detection surface of the detector 105.

[0047] Here, in observation state [4], a simulated representation is shown of a spherical sample 108 placed on the sample stage 103. However, the sample 108 is placed outside the field of view of observation. On the other hand, in observation state [5], unlike observation state [4], the sample 108 is placed within the field of view of observation. In observation state [4], since the sample 108 is placed outside the irradiation range of the electron beam, the shape of the surface of the sample stage 103 is observed based on the backscattered electrons 106 from the sample stage 103.

[0048] On the other hand, under observation conditions [5], sample 108 emits backscattered electrons 106 in response to electron beam irradiation. The trajectory of the backscattered electrons 106 is determined by the surface condition and tilt angle of sample 108. At the position shown in the figure, the electron beam is irradiated to the upper right part of sample 108, and the backscattered electrons 106 are distributed with high intensity in the upper right direction. As a result, the intensity of detector 105c, which is located on the right side of the detector group, is high, and the intensity of detector 105a, which is located on the left side, is low, and the difference ratio Dac is a value that depends on the surface tilt state of sample 108.

[0049] What is noteworthy here is that the distance d between the electron beam irradiation surface of sample 108 and the detection surface of detector 105 is shortened by the height of the sample. As mentioned in Figure 6, the value of the difference ratio Dac changes even on the same inclined surface when the distance d changes. As a result, an error occurs in the height measurement due to the difference in the difference in sample height, and consequently the difference in the difference ratio Dac due to the difference in distance d. In order to reduce this error, the second embodiment is characterized by performing a three-dimensional reconstruction calculation by applying conditions according to the height map of the sample for each region within the observation field of view.

[0050] Figure 10 is a flowchart showing an example of the processing content of a charged particle beam apparatus according to the second embodiment. In Figure 10, the charged particle beam apparatus starts electron beam irradiation of the sample 108 and performs predetermined image adjustments (step S101). Next, the charged particle beam apparatus detects backscattered electron (BSE) signals using a plurality of detectors 105 constituting a detector group, generates images for each of the plurality of detectors by image capture by the imaging device 110, and records imaging conditions including imaging WD as supplementary information to the image (step S102).

[0051] Next, the charged particle beam apparatus, specifically the arithmetic control unit 120 including a processor, derives a first calibration parameter corresponding to the imaging WD included in the supplementary information (step S103). Then, the arithmetic control unit 120 performs a three-dimensional reconstruction process using the derived first calibration parameter to generate a height map (step S104). In other words, the arithmetic control unit 120 generates a height map by provisionally generating a three-dimensional image of the sample based on the output of the detector group, i.e., the images for each of the multiple detectors generated in step S102, and the first calibration parameter derived in step S103. Up to this point, it is almost the same as Figure 7 described above.

[0052] Next, the arithmetic control unit 120 derives a second calibration parameter for each pixel, or for each pixel region consisting of multiple pixels, from the height map within the observation field obtained in step S104 (step S105). Finally, the arithmetic control unit 120 performs a three-dimensional reconstruction process using the second calibration parameter obtained in step S105 and displays the processing result on the display unit 125 (step S106). In other words, the arithmetic control unit 120 generates a three-dimensional image of the sample based on the output from the detector group, i.e., the images for each of the multiple detectors generated in step S102, and the second calibration parameter derived in step S105.

[0053] Here, the height map generated in step S104 reflects the first calibration parameter that compensates for the WD. Then, in step S105, a second calibration parameter is derived that compensates for the sample height in addition to the WD, based on the height map that reflects the first calibration parameter. Therefore, the calculation control unit 120 essentially generates a three-dimensional image of the sample based on the output from the detector group, i.e., the images for each of the multiple detectors generated in step S102, and the derived first and second calibration parameters. The calculation control unit 120 then displays the generated three-dimensional image of the sample on the 3D image display unit 1252.

[0054] The second calibration parameter is similarly determined from the relationship between distance d (or imaging WD) and difference ratio height, for example, as shown in Figure 8. However, while the first calibration parameter is derived using the same distance d (or imaging WD) value throughout the entire image, the second calibration parameter is derived individually for each pixel or pixel region within the image. That is, individual distance d (or imaging WD) values ​​are determined for each pixel or pixel region from the height map obtained in step S104, and individual calibration parameters are derived reflecting these individual distance d values.

[0055] The three-dimensional shape measurement results calculated using the first calibration parameter include errors due to differences in sample height, and consequently, differences in distance d, as mentioned above. On the other hand, the three-dimensional shape measurement results calculated using the second calibration parameter include only sufficiently small errors because they are further corrected based on sample height. This makes it possible to measure the three-dimensional shape with high accuracy even for samples with large height differences.

[0056] As mentioned above, the unit for deriving the second calibration parameter can be either per pixel or per pixel region. Here, a pixel region refers to an image within the observation field that has been divided into multiple sub-regions using its characteristics. For example, it may be a sub-region extracted by segmenting each object present in the sample by detecting the edges of the image. In this case, each pixel within each sub-region is considered to be at the same height. Derivation on a pixel-by-pixel basis allows for finer correction, but the computational load of the correction process increases, resulting in longer processing times. Correction on a pixel region basis allows for processing within the region using the same calibration parameter, thus reducing the computational load, but errors due to height differences within the region may occur.

[0057] Figure 10 illustrates the processing steps when a first calibration parameter corresponding to the working distance (WD) and a second calibration parameter corresponding to the sample height are combined. However, for example, when the working distance (WD) is fixed, in steps S103 and S104, the first calibration parameter may not be used, and a height map may simply be generated based on the images for each of the multiple detectors generated in step S102.

[0058] <Main effects of the second embodiment> As described above, the same effects as those described in the first embodiment can be obtained by using the method of the second embodiment. Furthermore, since it is possible to correct the difference ratio height error caused by the height of the sample, it is possible to perform three-dimensional shape measurement with minimal degradation of accuracy, even when measuring samples with large irregularities.

[0059] (Third embodiment) <Outline of a charged particle beam device> Figure 11 is a schematic diagram showing an example of the overall configuration of a charged particle beam apparatus according to the third embodiment. In Figure 11, the configuration related to the SEM is almost the same as the configuration shown in Figure 1. However, in Figure 11, an optical camera 109 is installed in the sample chamber 101. The optical camera 109 captures an optical image of the sample 108 and records the information of the optically captured image in the memory 151 in the optical imaging device 150. When an optical image is captured, the stage 104 moves the sample 108 to the position where the optical image is to be captured. The information processing unit 1221 in the calculation control unit 120 performs calculations on the captured optical image and analyzes the shape of the sample 108.

[0060] Several shape analysis methods are known for use in this process. For example, one method involves determining the height of feature points using the stereo method from multiple images of the tilted sample, or determining the focal point from multiple images with different focal lengths and then synthesizing them. The height maps obtained using these methods are used in place of the height map in step S104 in Figure 10. This allows for three-dimensional shape measurement by continuing with steps S105 and S106. The resolution of the optical camera 109 is generally lower than that of the SEM, but the height map required in step S104 only needs to have enough accuracy to derive calibration parameters. The optical camera 109 may also be installed separately from the sample chamber 101. In this case, an optical image should be captured before placing the sample 108 in the sample chamber 101, and then aligned with the SEM image.

[0061] <Measures to address sample height> Figure 12 is a flowchart showing an example of the processing steps of a charged particle beam apparatus according to the third embodiment. In Figure 12, the charged particle beam apparatus moves the sample 108 to the imaging position of the optical camera 109 (step S121). Next, the charged particle beam apparatus captures an optical image with the optical camera 109 (step S122). Subsequently, the charged particle beam apparatus, in particular the arithmetic control unit 120 including a processor, performs a three-dimensional reconstruction process on the acquired optical image, i.e., the captured image, using a suitable analysis method to generate a height map (step S123).

[0062] Next, the charged particle beam apparatus moves the sample 108 to the SEM observation position (step S124), starts irradiating the sample 108 with an electron beam, and performs predetermined image adjustments (step S125). Then, the charged particle beam apparatus detects backscattered electron (BSE) signals using multiple detectors 105 that constitute the detector group, and generates images for each of the multiple detectors by image capture by the imaging device 110 (step S126).

[0063] Next, the charged particle beam apparatus, specifically the calculation control unit 120, derives calibration parameters for each pixel or pixel region from the height map obtained in step S123 (step S127). Then, the calculation control unit 120 performs a three-dimensional reconstruction process using the calibration parameters obtained in step S127 and displays the processing results on the display unit 125 (step S128).

[0064] <Main effects of the third embodiment> As described above, the method of the third embodiment can be used to obtain the same effects as those described in the first and second embodiments. Furthermore, it is possible to perform highly accurate three-dimensional shape measurements even for samples that include steeply sloped or shielded areas, where accuracy cannot be obtained or artifacts occur with the three-dimensional shape measurement using multiple detectors arranged in different directions as described above.

[0065] (Fourth embodiment) <Measures to address sample height> The method for obtaining a height map is not limited to the above; various methods are possible. In the third embodiment described above, an optical camera was newly installed, but for example, a method that utilizes the normal functions of the SEM by electron beam irradiation without adding dedicated hardware may be used. In the fourth embodiment, a method for obtaining a height map by utilizing the normal functions of the SEM will be described. The charged particle beam apparatus according to the fourth embodiment has the configuration shown in Figure 1, for example.

[0066] Figure 13 is a schematic diagram showing an example of the shape of a sample to be measured in the fourth embodiment. In Figure 13, the upper panel shows a plan view of the sample, in other words, an SEM image in the observation field, and the lower panel shows a cross-sectional view of the sample viewed from the side. Here, a sample to be inspected for foreign matter is simulated, and an example is shown in which two foreign matter particles 302 and 303 have been extracted onto a substrate part 301 such as a filter for filtering and extracting foreign matter. In foreign matter inspection, morphological information (size such as particle size, degree of irregularity, etc.) and elemental information of the extracted foreign matter are required, but in addition, morphological information regarding the height of the foreign matter is also an important monitoring indicator. For this reason, it is desirable to obtain the height information of the foreign matter with high accuracy.

[0067] Figure 14 is a flowchart showing an example of the processing steps when performing three-dimensional reconstruction in a charged particle beam apparatus according to the fourth embodiment. In Figure 14, the charged particle beam apparatus starts electron beam irradiation of the sample and performs predetermined image adjustments (step S141). Next, the charged particle beam apparatus detects a signal using the detector 105 and generates an SEM image by image capture with the imaging device 110 (step S142). At this time, it is desirable that the detector 105 acquires an image that facilitates segmentation, which will be described later. Examples of methods for acquiring such an image include using a secondary electron detector as the detector 105, or using a composite image of BSE signals from multiple detectors.

[0068] Next, the charged particle beam apparatus, specifically the arithmetic control unit 120 including a processor, performs region segmentation on the acquired SEM image by applying edge detection processing according to its features (step S143). As a result, for example in Figure 13, a segmented region corresponding to foreign object 302 and another segmented region corresponding to foreign object 303 are formed. Subsequently, the charged particle beam apparatus searches for a focal point for each segmented region by performing autofocus for each segmented region. Based on the focal point, the charged particle beam apparatus determines the working distance (WD) value for each segmented region and generates a height map for each segmented region (step S144).

[0069] Next, the calculation control unit 120 derives calibration parameters for each divided region from the obtained height map (step S145). Then, the calculation control unit 120 performs a three-dimensional reconstruction process using the obtained calibration parameters and displays the processing result on the display unit 125 (step S146). Note that the height map obtained in step S144 has a low resolution because each divided region has a certain height, but as mentioned above, the detailed shape of the foreign object is not necessary for foreign object inspection; knowing the height is sufficient.

[0070] <Main effects of the fourth embodiment> As described above, the method of the fourth embodiment can be used to obtain the same effects as those described in the first and second embodiments. Furthermore, unlike the method of the third embodiment, it is possible to measure the height of foreign objects with high precision without providing special hardware, and thus without increasing costs.

[0071] While several embodiments have been described in detail, it is clear that further accuracy can be improved by combining these embodiments. Furthermore, although the embodiments described used a ring-type backscattered electron detector of a scanning electron microscope (SEM) as an example, similar effects can be obtained by applying the method to three-dimensional reconstruction processing using multiple other detectors, or by applying it to other charged particle beam devices such as FIB-SEM.

[0072] The present invention has been described in detail above based on embodiments, but the present invention is not limited to the embodiments described above and can be modified in various ways without departing from its essence. For example, the embodiments described above are described in detail in order to explain the present invention in an easy-to-understand manner and are not necessarily limited to those having all the described configurations. Furthermore, it is possible to replace a part of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add a configuration from another embodiment to the configuration of one embodiment. In addition, it is possible to add, delete, or replace a part of the configuration of each embodiment with a configuration from another embodiment. [Explanation of Symbols]

[0073] 101: Sample Room 102: Electron optical system 103: Sample stage 104: Stage 105: Detector 106: Backscattered electron 109: Optical camera 108: Sample 110: Charged particle beam imaging device 111: Memory 120: Arithmetic Control Unit 121:CPU 122: Main memory 1221: Information Processing Unit 1224: Display Processing Unit 123:Secondary storage device 124: Input section 125: Display section 1251: Observation Image Display Unit 1252: 3D Image Display Unit 130: Stage control device 131: Coordinate memory unit 132: Control Unit 133: Drive unit 150: Optical imaging device 151: Memory 161: Electron source 162: Condenser lens 163: Deflector 164: Objective lens 170: Optical System Control Device 171: Memory 301: Base material part 302: Foreign object 303: Foreign object 501: Cross-sectional profile (solid line) 502: Wire cross-section simulated circle (dotted line) 503: Cross-sectional profile with errors (dotted line)

Claims

1. A detector group consisting of multiple detectors arranged in different directions from each other detects that detect charged particles emitted from the sample, Processor and It has, The aforementioned processor, A first calibration parameter is derived according to the working distance, which is based on the distance between the sample and the detector group when detecting the charged particles. Based on the height map of the sample generated in advance, a second calibration parameter is derived that reflects the individual distance between the sample and the detector group for each pixel or pixel region within the observation field of view of the detector group, for each pixel or pixel region. A three-dimensional image of the sample is generated based on the output from the detector group and the derived first and second calibration parameters. Charged particle beam device.

2. In the charged particle beam apparatus according to claim 1, The aforementioned group of detectors detects backscattered electrons, The plurality of detectors are arranged symmetrically with respect to the electron beam irradiation axis onto the sample. Charged particle beam device.

3. A detector group consisting of multiple detectors arranged in different directions from each other detects that detect charged particles emitted from the sample, Processor and It has, The aforementioned processor, Based on the height map of the sample generated in advance, calibration parameters are derived that reflect the individual distance between the sample and the detector group for each pixel or pixel region within the observation field of view of the detector group, for each pixel or pixel region. A three-dimensional image of the sample is generated based on the output from the detector group and the derived calibration parameters. Charged particle beam device.

4. In the charged particle beam apparatus according to claim 3, The processor generates a height map of the sample by provisionally generating a three-dimensional image of the sample based on the output from the detector group. Charged particle beam device.

5. In the charged particle beam apparatus according to claim 3, The height map of the sample is generated using an image of the sample taken by an optical camera. Charged particle beam device.

6. In the charged particle beam apparatus according to claim 3, The observation field of view of the detector group is divided into multiple divided regions, and a height map of the sample is generated by searching for a focal point in each of the divided regions. Charged particle beam device.

7. In the charged particle beam apparatus according to any one of claims 3 to 6, The aforementioned group of detectors detects backscattered electrons, The plurality of detectors are arranged symmetrically with respect to the electron beam irradiation axis onto the sample. Charged particle beam device.

8. A detector group consisting of multiple detectors arranged in different directions from each other detects that detect charged particles emitted from the sample, Processor and A three-dimensional shape measurement method using a charged particle beam apparatus equipped with, A first calibration parameter is derived according to the working distance, which is based on the distance between the sample and the detector group when detecting the charged particles. Based on the height map of the sample generated in advance, a second calibration parameter is derived that reflects the individual distance between the sample and the detector group for each pixel or pixel region within the observation field of view of the detector group, for each pixel or pixel region. A three-dimensional image of the sample is generated based on the output from the detector group and the derived first and second calibration parameters. Three-dimensional shape measurement method.

9. In the three-dimensional shape measurement method according to claim 8, A height map of the sample is generated by provisionally generating a three-dimensional image of the sample based on the output from the detector group and the first calibration parameter. A three-dimensional image of the sample is generated based on the output from the detector group and the second calibration parameter derived from the height map of the sample. Three-dimensional shape measurement method.

10. In the three-dimensional shape measurement method according to claim 8, A height map of the sample is generated using an image of the sample captured by an optical camera. Three-dimensional shape measurement method.

11. In the three-dimensional shape measurement method according to claim 8, The observation field of view of the detector group is divided into multiple divided regions, and a height map of the sample is generated by searching for a focal point in each of the divided regions. Three-dimensional shape measurement method.

12. In the three-dimensional shape measurement method according to any one of claims 8 to 11, The aforementioned group of detectors detects backscattered electrons, The plurality of detectors are arranged symmetrically with respect to the electron beam irradiation axis onto the sample. As output from the detector group, a three-dimensional image of the sample is generated using the image of the backscattered electrons detected by the detector group. Three-dimensional shape measurement method.

Images