A method for calibrating the magnification of a scanning electron microscope

Abstract

This application relates to the field of semiconductor technology and provides a method for magnification calibration of a scanning electron microscope (SEM). The method includes: in TEM imaging mode, determining TEM measurement data of key structures of a TEM sample, wherein the TEM sample is a TEM sample with known process structure dimensions or clear lattice fringes; in SEM imaging mode, determining SEM measurement datasets of the key structures of the TEM sample at different SEM magnifications; and calibrating various SEM magnifications of the scanning electron microscope based on the TEM measurement data and the SEM measurement datasets. This application is simple to operate, does not rely on standard samples, has lower calibration costs, and provides higher calibration accuracy.

Description

Technical Field

[0001] This application relates to the field of semiconductor technology, and in particular to a method for calibrating the magnification of a scanning electron microscope. Background Technology

[0002] In related technologies, magnification calibration of scanning electron microscopy (SEM) typically relies on standard samples (such as... Figure 1 (As shown). However, there is an uncertain error between the true value (i.e., the actual size) of the structure of the standard sample (hereinafter referred to as "standard sample") and the process reference value. Moreover, during SEM characterization, its structural pattern is easily affected by factors such as charge and angular contrast, leading to uncertain errors between the measured value and the true value. This directly affects the calibration accuracy of SEM magnification. In addition, standard samples are relatively expensive and have a short service life (generally 3 years). If not stored properly, the accuracy of the standard sample will be greatly reduced, requiring repeated purchases of new standard samples, resulting in high calibration costs.

[0003] It is evident that traditional SEM magnification calibration methods rely on standard samples, resulting in high calibration costs and low calibration accuracy. Summary of the Invention

[0004] In view of this, this application provides a method for magnification calibration of scanning electron microscopes, which aims to solve the problems of traditional SEM magnification calibration methods that rely on standard samples, have high calibration costs, and low calibration accuracy.

[0005] To achieve the above-mentioned objectives, embodiments of this application provide a method for magnification calibration of a scanning electron microscope, comprising:

[0006] In TEM imaging mode, TEM measurement data are used to determine the key structures of a TEM sample, wherein the TEM sample is a TEM sample with known process structure dimensions or with clear lattice fringes;

[0007] In SEM imaging mode, SEM measurement datasets of the key structures of the TEM sample at different SEM magnifications were determined;

[0008] Based on the TEM measurement data and SEM measurement dataset, the various SEM magnifications of the scanning electron microscope are calibrated.

[0009] Compared with existing technologies, the beneficial effects of this application include at least the following: in TEM imaging mode, determining TEM measurement data of key structures of TEM samples with known process structure dimensions or clear lattice fringes; in SEM imaging mode, determining SEM measurement datasets of the key structures of the aforementioned TEM samples at different SEM magnifications; and calibrating various SEM magnifications of the scanning electron microscope based on the TEM measurement data and the SEM measurement datasets. This calibration method is simple to operate, does not rely on standard samples, has lower calibration costs, and higher calibration accuracy. Attached Figure Description

[0010] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 This is a schematic diagram of the standard sample used in traditional SEM magnification calibration methods;

[0012] Figure 2 This is a schematic flowchart of a magnification calibration method for a scanning electron microscope provided in an embodiment of this application;

[0013] Figure 3 This is a method provided in this application for marking structural feature points of key structures in the first image region of a target TEM image of a TEM sample;

[0014] Figure 4 This is one method provided in this application for marking structural feature points of key structures in the second image region of the first SEM image of a TEM sample. Detailed Implementation

[0015] To make the technical problems, technical solutions, and beneficial effects to be solved by this application clearer, the following detailed description is provided in conjunction with embodiments. It should be understood that the specific embodiments described herein are only for explaining this application, but the implementation of this application is not limited thereto.

[0016] Magnification calibration of scanning electron microscopes is crucial for measurement accuracy. It is the core basis for calculating microscopic dimensions, and lack of calibration can lead to data deviations, potentially causing product problems or misjudging the relationship between material structure and properties in fields such as precision manufacturing and materials research. Simultaneously, calibration ensures the repeatability and comparability of experimental results, preventing inconsistencies in repeated experiments using the same equipment due to parameter drift. It also ensures that measurement results from different devices or laboratories are based on a unified standard, guaranteeing the validity of data in scientific collaborations.

[0017] However, traditional SEM magnification calibration methods rely on standard samples, resulting in high calibration costs and low calibration accuracy.

[0018] A transmission electron microscope (TEM) is an instrument widely used for semiconductor structure testing. It has ultra-high resolution (atomic-level resolution), with an accuracy of less than 0.1 nm. It can directly observe the atomic arrangement, lattice structure, and even individual atoms of materials. Therefore, the dimensional data measured by it can be used as a natural (absolute) scale. Combined with SEM-TEM technology, standard-free calibration of SEM magnification can be achieved, thereby reducing calibration costs and improving calibration accuracy.

[0019] Therefore, this application provides a magnification calibration method for scanning electron microscopes. Utilizing the lattice or nanostructure observed by TEM as an absolute scale, the method measures TEM measurement data of key structures of the TEM sample in TEM imaging mode and SEM measurement datasets of key structures at different SEM magnifications in SEM imaging mode through sample positioning and image matching. Based on the TEM and SEM measurement datasets, each SEM magnification is calibrated, achieving high-precision magnification calibration. This method does not rely on standard samples, avoiding the impact of uncertainties between the true structural value and process reference value of the standard sample, as well as uncertainties between the measured value and the true value, on calibration accuracy. This improves calibration accuracy and provides a reliable and universal solution for nanoscale metrology.

[0020] The true structural value of the standard sample provides a traceability basis for its process reference value. The process reference value provides a calibration benchmark for the measuring equipment, and the measured values ​​are fed back into the process optimization through error analysis, with the aim of approximating the true structural value of the standard sample as closely as possible.

[0021] Figure 2 This is a schematic flowchart illustrating a magnification calibration method for a scanning electron microscope according to an embodiment of this application. Please refer to... Figure 2 This application provides a method for calibrating the magnification of a scanning electron microscope, including:

[0022] S201. In TEM imaging mode, determine the TEM measurement data of the key structure of the TEM sample, wherein the TEM sample is a TEM sample with known process structure size or with clear lattice fringes.

[0023] The critical structure (CT) usually refers to the through-hole structure in a chip.

[0024] The through-hole structure facilitates measurement, has a larger aspect ratio, and makes it easier to obtain key and accurate dimensional data. It also makes it easier to locate the same area (the area where the key structure is located) in both TEM and SEM imaging modes. This helps to avoid introducing large measurement errors in both TEM images containing the key structure and SEM images containing the key structure.

[0025] TEM measurement data refers to the actual dimensions of key structures of a TEM sample under TEM imaging mode.

[0026] Lattice fringes are periodic bright and dark stripes that appear in TEM (especially HRTEM (high resolution transmission electron microscopy)) images. They are essentially characteristic signals formed by the interaction between the "crystal planes" in the material's crystal structure and the electron beam, and directly correspond to the distance between the crystal planes (lattice spacing).

[0027] Using TEM samples with clear lattice fringes as standard samples can ensure the "structural determinism" and "signal stability" of the standard sample itself, which is beneficial to ensuring the accuracy of subsequent calibration of various SEM magnifications of scanning electron microscopes using the standard sample.

[0028] S202. In SEM imaging mode, determine the SEM measurement dataset of the key structure of the TEM sample at different SEM magnifications.

[0029] SEM magnification typically includes: low magnification (5×~100×), medium magnification (100×~10000×), and high magnification (10000×~1000000×).

[0030] S203. Based on the TEM measurement data and SEM measurement dataset, calibrate each SEM magnification of the scanning electron microscope.

[0031] Compared to traditional SEM magnification calibration methods, the technical solution provided in this application utilizes the lattice or nanostructure observed by TEM as an absolute benchmark. Through sample positioning and image matching, it measures TEM measurement data of key structures of the TEM sample in TEM imaging mode and SEM measurement datasets of key structures of the TEM sample at different SEM magnifications in SEM imaging mode. Based on the TEM measurement data and SEM measurement datasets, the scanning electron microscope's various SEM magnifications are calibrated, achieving high-precision magnification calibration. This method does not rely on standard samples, avoiding the impact of uncertainties between the true structural value and process reference value of the standard sample, as well as uncertainties between the measured value and the true value, on calibration accuracy. This improves calibration accuracy and provides a reliable and universal solution for nanoscale metrology.

[0032] In some implementations, TEM measurement data that determine the key structures of a TEM sample include:

[0033] Obtain the target TEM image of the TEM sample at the target TEM magnification;

[0034] In the target TEM image, the first image region corresponding to the key structure of the TEM sample is located, and based on the first image feature information of the first image region, the TEM measurement data of the key structure of the TEM sample at the target TEM magnification is determined.

[0035] The typical range of TEM magnification is 20× to 2,000,000×. The target TEM magnification can be any value between 20× and 2,000,000×.

[0036] The first image feature information usually refers to the location of feature points of the key structures of the TEM sample in the first image region.

[0037] As an example, firstly, the TEM sample is placed under a TEM lens, and the magnification of the TEM lens is adjusted to the target TEM magnification. A target TEM image of the TEM sample at the target TEM magnification is then acquired. Next, the first image region corresponding to the key structure of the TEM sample is located in the target TEM image, and the feature points of this key structure (usually the corner points of a through-hole structure) are marked. Figure 3As shown, assuming the first image region of the target TEM image of the TEM sample includes three through-hole structures (CT), the corner points of these three through-hole structures (such as points A1, A2, A3, and A4) can be marked in the first image region. Next, based on the structural feature points of the key structures marked in the first image region, the first distance d1 between points A1 and A3, and the second distance d2 between points A2 and A4 are measured using a ruler. Then, based on the first distance d1, the second distance d2, and the target TEM magnification, the actual size (true size) of the key structures of the TEM sample at the target TEM magnification is calculated, thus obtaining the TEM measurement data.

[0038] For example, the actual size of the key structure of the TEM sample at the target TEM magnification can be calculated according to Equation (1).

[0039] (1);

[0040] In equation (1), This indicates the actual size of the key structures of the TEM sample at the target TEM magnification. The first distance value between structural feature point A1 and point A3, representing the key structure marked in the first image region; The second distance value represents the structural feature point A2 and point A4 of the key structure marked in the first image region; Indicates the target TEM magnification.

[0041] In some implementations, the SEM measurement dataset includes multiple first SEM measurement data, and one first SEM measurement data corresponds to one first SEM magnification.

[0042] The SEM measurement datasets for the key structures of the TEM samples at different SEM magnifications were determined, including:

[0043] If the TEM sample is a conductor sample, then under the first measurement conditions, a first SEM image set of the TEM sample at different SEM magnifications is acquired. The first SEM image set includes multiple first SEM images, and one first SEM image corresponds to one first SEM magnification. The first measurement conditions include: a first accelerating voltage of 5~15 kV and a first scanning speed of 10~30 seconds / frame.

[0044] For each first SEM image, find the second image region in the first SEM image that corresponds to the key structure of the TEM sample;

[0045] Based on the second image feature information corresponding to the second image region, the first SEM measurement data of the key structure of the TEM sample at the first SEM magnification corresponding to the first SEM image is determined.

[0046] The second image feature information usually refers to the location of feature points of key structures of the TEM sample in the second image region.

[0047] As an example, assume that the SEM measurement dataset includes first SEM measurement data 01 (corresponding to SEM low magnification), first SEM measurement data 02 (corresponding to SEM medium magnification), and first SEM measurement data 03 (corresponding to SEM high magnification).

[0048] At low SEM magnification, the working distance (WD) is set to short WD, the first accelerating voltage is 5~15 kV, and the first scanning speed is 10~30 seconds / frame. The magnification error of the X and Y axes in the SEM imaging mode is calibrated to be less than a preset error threshold (typically 1%) using a standard grating. The TEM sample is then scanned under the low-magnification SEM lens to obtain the first SEM image 01 corresponding to the low SEM magnification. The short WD value is typically 3~4 mm.

[0049] Similarly, referring to the steps described above for obtaining the first SEM image 01 at a low SEM magnification, the first SEM image 02 at a medium SEM magnification is obtained.

[0050] At high SEM magnification, the working distance (WD) is set to a length WD, the first accelerating voltage is 5~15 kV, and the magnification error of the X and Y axes in the SEM imaging mode is calibrated to be less than a preset error threshold (generally 1%) using a standard grating. The TEM sample is placed under the high-magnification SEM lens for slow scanning to obtain the first SEM image 03 corresponding to the high SEM magnification. The length WD is generally in the range of 4~5 mm.

[0051] Setting the working distance (WD) to short WD at low magnification of SEM and setting the working distance (WD) to long WD at high magnification of SEM can effectively prevent magnification drift caused by changes in focal length.

[0052] By controlling the first accelerating voltage within the range of 5~15kV, the resolution and penetration depth can be balanced, which is beneficial to improving image quality.

[0053] At low and medium magnification SEM magnification, controlling the first scan speed to 10-30 seconds / frame and using a slow scan mode at high SEM magnification (greater than 50000×) can reduce electron beam drift.

[0054] In some implementations, based on the second image feature information corresponding to the second image region, the first SEM measurement data of the key structure of the TEM sample at the first SEM magnification corresponding to the first SEM image is determined, including:

[0055] Mark at least two structural feature points corresponding to the key structures of the TEM sample in the second image region;

[0056] Based on at least two of the structural feature points, the first SEM measurement data of the key structure of the TEM sample at the first SEM magnification corresponding to the first SEM image is determined.

[0057] For ease of understanding, please continue using the above example. Figure 4 Cross-shaped grooves can be used to mark the locations of at least two structural feature points corresponding to the key structures of the TEM sample in the second image region, such as... Figure 4 Points A1, A2, A3, and A4 are selected. Then, the third distance value d3 between point A1 and point A3, and the fourth distance value d4 between point A2 and point A4 are measured using a ruler. Based on the third distance value d3, the fourth distance value d4, and the first SEM magnification, the actual size (true size) of the key structure of the TEM sample under the first SEM magnification is calculated, thus obtaining the first SEM measurement data O1.

[0058] For example, the actual size of the key structure of the TEM sample at the first SEM magnification can be calculated according to Equation (2).

[0059] (2);

[0060] In equation (2), This indicates the actual size of the key structure of the TEM sample at the first SEM magnification (low SEM magnification); The third distance value represents the structural feature point A1 and point A3 of the key structure marked in the second image region; The fourth distance value represents the structural feature point A2 and point A4 of the key structure marked in the second image region; This indicates the first SEM magnification (low SEM magnification).

[0061] Similarly, the calculation steps for the first SEM measurement data 01 can be referred to above to calculate the first SEM measurement data 02 and the first SEM measurement data 03, which will not be repeated here.

[0062] Typically, through-hole structures with larger structural dimensions in TEM samples are selected as key structures. The corners of through-hole structures are easy to measure, which can avoid introducing large measurement errors in TEM and SEM images, and is beneficial to improving the reliability and accuracy of the calibration results of the magnification of scanning electron microscopes.

[0063] In some implementations, the SEM measurement dataset includes multiple second SEM measurement data, with one second SEM measurement data corresponding to one second SEM magnification.

[0064] The SEM measurement datasets for the key structures of the TEM samples at different SEM magnifications were determined, including:

[0065] If the TEM sample is an insulator sample, a conductive layer is prepared on the surface of the insulator sample to obtain a coated sample.

[0066] Under the second measurement conditions, a second SEM image set of the coating sample at different SEM magnifications is acquired. The second SEM image set includes multiple second SEM images, and one second SEM image corresponds to one second SEM magnification. The second measurement conditions include: a second accelerating voltage of 1~3kV and a second scanning speed of 10~30 seconds / frame.

[0067] For each second SEM image, a third image region corresponding to the key structure of the TEM sample is located in the second SEM image. Based on the third image feature information corresponding to the third image region, the second SEM measurement data of the key structure of the TEM sample at the second SEM magnification corresponding to the second SEM image is determined.

[0068] As an example, a conductive layer can be prepared on the surface of the insulator sample using existing plating processes to obtain a plated sample.

[0069] By preparing a conductive layer on the surface of an insulator sample, image stretching caused by the charging effect can be suppressed, thereby ensuring image quality and the accuracy of subsequent calculations.

[0070] As an example, assume that the SEM measurement dataset includes second SEM measurement data 01 (corresponding to low SEM magnification), second SEM measurement data 02 (corresponding to medium SEM magnification), and second SEM measurement data 03 (corresponding to high SEM magnification).

[0071] Similarly, the calculation steps for the first SEM measurement data 01 (corresponding to low SEM magnification), the first SEM measurement data 02 (corresponding to medium SEM magnification), and the first SEM measurement data 03 (corresponding to high SEM magnification) can be referred to above to calculate the second SEM measurement data 01 (corresponding to low SEM magnification), the second SEM measurement data 02 (corresponding to medium SEM magnification), and the second SEM measurement data 03 (corresponding to high SEM magnification), which will not be repeated here.

[0072] In some embodiments, before acquiring a second set of SEM images of the coating sample at different SEM magnifications under the second measurement conditions, the method further includes:

[0073] The magnification error of the X and Y axes in SEM imaging mode is less than the preset error threshold when calibrated with a standard grating.

[0074] The preset error threshold can be set according to the actual situation, and is usually set to 1%.

[0075] As an example, the specific procedure for calibrating the X-axis and Y-axis magnification errors in SEM imaging mode using a standard grating to ensure they are less than a preset error threshold is as follows:

[0076] Step 1: Preparation. Mount the TEM sample onto a standard grating with a known precise spacing (e.g., 1000 nm), turn on the microscope and warm it up for 30 minutes to stabilize it.

[0077] The second step is to take a good photo. First, find a clean and flat area of ​​the raster at low magnification. Then, at the target calibration magnification, strictly focus and correct astigmatism to ensure the sharpest and clearest image.

[0078] Step 3: Measurement and Calculation. First, measure the spacing between multiple grid cells horizontally, and calculate the average value of the measured grid spacing in the X direction (X-axis). M_x ; Measure the spacing between multiple grid cells vertically, and calculate the average value of the measured grid spacing in the Y direction (Y-axis). M_y Then, calculate the magnification error of the X-axis according to formula (3); calculate the magnification error of the Y-axis according to formula (4).

[0079] (3);

[0080] In equation (3), The average value of the measured spacing of multiple grids in the X direction (X-axis); Standard values ​​representing the spacing between multiple grid cells in the X direction (X-axis); This indicates the magnification error of the X-axis.

[0081] (4);

[0082] In equation (4), This represents the average measured spacing of multiple grid cells in the Y direction (Y-axis). Standard values ​​representing the spacing between multiple grid cells in the Y direction (Y-axis); This indicates the scaling error along the Y-axis.

[0083] Step 4: Calibration and Verification. If the magnification error of the X-axis and Y-axis is greater than or equal to the preset error threshold, enter the calibration menu of the SEM software, input the standard value, and repeat the measurement as prompted, then proceed to Step 5.

[0084] Step 5: Verify again. A new photo must be taken and the measurements and calculations repeated until the magnification error of the X-axis and Y-axis is less than the preset threshold.

[0085] In some implementations, based on the TEM measurement data and SEM measurement dataset, the various SEM magnifications of the scanning electron microscope are calibrated, including:

[0086] For each first SEM magnification, the first SEM magnification is calibrated based on the TEM measurement data and the first SEM measurement data.

[0087] As an example, assuming the SEM measurement dataset includes first SEM measurement data 01 (corresponding to SEM low magnification), first SEM measurement data 02 (corresponding to SEM medium magnification), and first SEM measurement data 03 (corresponding to SEM high magnification), then the SEM low magnification can be calibrated based on the first SEM measurement data 01 and the TEM measurement data; the SEM medium magnification can be calibrated based on the first SEM measurement data 02 and the TEM measurement data; and the SEM high magnification can be calibrated based on the first SEM measurement data 03 and the TEM measurement data.

[0088] In some implementations, calibrating the first SEM magnification based on the TEM measurement data and the first SEM measurement data includes:

[0089] Calculate the measurement error value between the TEM measurement data and the first SEM measurement data;

[0090] If the measurement error value is less than or equal to a preset threshold, then the first SEM magnification is marked as the correct SEM magnification.

[0091] The preset threshold can be set according to the actual situation, and is usually set to 1%.

[0092] As an example, for SEM low magnification, first calculate the measurement error value 01 between the TEM measurement data and the first SEM measurement data 01; if the measurement error value 01 is less than or equal to a preset threshold (such as 1%), then the SEM low magnification is marked as the correct SEM magnification, and there is no need to calibrate the SEM low magnification.

[0093] Similarly, the above method can be used to determine whether the medium-magnification and high-magnification of SEM need to be accurate.

[0094] In some implementations, after calculating the measurement error value between the TEM measurement data and the first SEM measurement data, the method further includes:

[0095] If the measurement error value is greater than a preset threshold, the first SEM magnification is marked as the SEM magnification to be calibrated.

[0096] Obtain target image feature information of the key structure of the TEM sample in the target image region at the magnification of the SEM to be calibrated;

[0097] Based on the target image feature information and TEM measurement data, the magnification of the SEM to be calibrated is calibrated.

[0098] Target image feature information typically refers to the number of pixels (i.e., resolution) and pixel size (the actual physical size occupied by a single pixel on a sensor or display) of key structures of the TEM sample in the target image region.

[0099] In some implementations, the magnification of the SEM to be calibrated is calibrated based on the target image feature information and TEM measurement data, including:

[0100] Based on the target image feature information and TEM measurement data, the actual SEM magnification is calculated;

[0101] The actual SEM magnification is used to correct the SEM magnification to be calibrated.

[0102] For ease of understanding, let's continue with the example above. If the measurement error value 01 is greater than the preset threshold (e.g., 1%), then the low magnification of the SEM will be marked as the SEM magnification to be calibrated.

[0103] For example, the actual SEM magnification can be calculated according to equation (3).

[0104] (5);

[0105] In equation (5), This indicates the actual SEM magnification. This indicates the actual size of the key structures of the TEM sample at the target TEM magnification. P represents the number of pixels of the key structure of the TEM sample in the target image region; P represents the pixel size of the key structure of the TEM sample in the target image region.

[0106] Finally, the SEM magnification to be calibrated is corrected to the actual SEM magnification.

[0107] In summary, the magnification calibration method for scanning electron microscopes provided in this application utilizes the lattice or nanostructure observed by TEM as an absolute scale. It eliminates the need for standard samples, avoiding the influence of uncertainties between the true structural value and the process reference value of the standard sample, as well as the uncertainties between the measured value and the true value, on calibration accuracy. This improves calibration accuracy to below ±1%. This calibration method is applicable to high-resolution SEM, in-situ experiments, and nanomaterial research, with a wide range of applications. This method not only improves the reliability of measurements but also establishes a new technical path for standardized nanoscale characterization, and helps improve the accuracy and repeatability of microstructure research, facilitating widespread application.

[0108] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method of magnification calibration of a scanning electron microscope, characterized by, The application relates to a method for calibrating scanning electron microscope (SEM) magnification, and belongs to the technical field of electron microscopy. In a TEM imaging mode, TEM measurement data of a key structure of a TEM sample are determined, the TEM sample being a TEM sample with a known process structure size or clear lattice fringes; the key structure refers to a via structure in a chip; In an SEM imaging mode, a SEM measurement data set of the key structure of the TEM sample under different SEM magnifications is determined; the SEM measurement data set comprises a plurality of first SEM measurement data, one first SEM measurement data corresponding to one first SEM magnification; Based on the TEM measurement data and the SEM measurement data set, each SEM magnification of a scanning electron microscope is calibrated; The TEM measurement data of the key structure of the TEM sample are determined, comprising: A target TEM image of the TEM sample under a target TEM magnification is acquired; In the target TEM image, a first image region corresponding to the key structure of the TEM sample is found, and based on first image feature information of the first image region, TEM measurement data of the key structure of the TEM sample under the target TEM magnification are determined; the first image feature information refers to the positions of feature points of the key structure of the TEM sample in the first image region; The SEM measurement data set of the key structure of the TEM sample under different SEM magnifications is determined, comprising: If the TEM sample is a conductor sample, a first SEM image set of the TEM sample under different SEM magnifications is acquired under a first measurement condition, the first SEM image set comprising a plurality of first SEM images, one first SEM image corresponding to one first SEM magnification; For each first SEM image, a second image region corresponding to the key structure of the TEM sample in the first SEM image is found; Based on second image feature information corresponding to the second image region, first SEM measurement data of the key structure of the TEM sample under the first SEM magnification corresponding to the first SEM image are determined; the second image feature information refers to the positions of feature points of the key structure of the TEM sample in the second image region; Based on the TEM measurement data and the SEM measurement data set, each SEM magnification of a scanning electron microscope is calibrated, comprising: For each first SEM magnification, a measurement error value between the TEM measurement data and the first SEM measurement data is calculated; If the measurement error value is greater than a preset threshold value, the first SEM magnification is marked as a to-be-calibrated SEM magnification; Target image feature information of a target image region of the key structure of the TEM sample under the to-be-calibrated SEM magnification is acquired; calculating an actual SEM magnification based on the target image feature information and the TEM measurement data; the actual SEM magnification is calculated according to the following formula: wherein, represents the actual SEM magnification; represents the actual size of the key structure of the TEM sample at the target TEM magnification; represents the pixel number of the key structure of the TEM sample in the target image area; P represents the pixel size of the key structure of the TEM sample in the target image area; The to-be-calibrated SEM magnification is corrected to the actual SEM magnification.

2. The magnification calibration method of a scanning electron microscope according to claim 1, wherein The first measurement condition comprises that a first acceleration voltage is 5-15 kV and a first scanning speed is 10-30 seconds / frame.

3. The magnification calibration method of a scanning electron microscope according to claim 1, wherein determining, based on second image feature information corresponding to the second image region, first SEM measurement data of the key structure of the TEM sample at a first SEM magnification corresponding to the first SEM image, including: labeling at least two structure feature points corresponding to the key structure of the TEM sample in the second image region; determining, based on the at least two structure feature points, first SEM measurement data of the key structure of the TEM sample at a first SEM magnification corresponding to the first SEM image.

4. The magnification calibration method of a scanning electron microscope according to claim 1, wherein The SEM measurement data set includes a plurality of second SEM measurement data, and one second SEM measurement data corresponds to one second SEM magnification. determining a SEM measurement data set of the key structure of the TEM sample at different SEM magnifications, including: if the TEM sample is an insulator sample, preparing a conductive layer on the surface of the insulator sample to obtain a plated sample; under second measurement conditions, obtaining a second SEM image set of the plated sample at different SEM magnifications, the second SEM image set including a plurality of second SEM images, and one second SEM image corresponding to one second SEM magnification; the second measurement conditions include: a second acceleration voltage of 1-3kV, and a second scanning speed of 10-30 seconds / frame; for each second SEM image, finding a third image region in the second SEM image corresponding to the key structure of the TEM sample, and determining, based on third image feature information corresponding to the third image region, second SEM measurement data of the key structure of the TEM sample at a second SEM magnification corresponding to the second SEM image.

5. The magnification calibration method of a scanning electron microscope according to claim 4, wherein Before obtaining the second SEM image set of the plated sample at different SEM magnifications under the second measurement conditions, the method further includes: calibrating the magnification error of the X-axis and the Y-axis in the SEM imaging mode to be less than a preset error threshold through a standard grating.

6. The method of magnification calibration for a scanning electron microscope of claim 1, wherein, According to the TEM measurement data and the first SEM measurement data, calibrating the first SEM magnification, including: calculating a measurement error value between the TEM measurement data and the first SEM measurement data; if the measurement error value is less than or equal to a preset threshold, marking the first SEM magnification as a correct SEM magnification.

Images