A CD-SEM measurement method and device for semiconductor device layout graphics

Through the method of automatically calculating geometric features and generating measurement parameters, the time-consuming problem of CD-SEM image data collection is solved, a fast and accurate measurement process is achieved, and semiconductor device manufacturing and R&D efficiency is improved.

CN118501504BActive Publication Date: 2024-11-01ZHEJIANG UNIV +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410958466.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-07-17
Publication Date
2024-11-01
Estimated Expiration
2044-07-17

AI Technical Summary

Technical Problem

In the prior art, CD-SEM image data collection consumes a lot of manual and machine time, and relies on engineer experience, resulting in an extended semiconductor device manufacturing and R&D cycle.

Method used

By establishing measurement graphic layout files and measurement information parameters of CD-SEM machines, using ray method and Fourier description operators to calculate geometric features, automatically determine the types and parameters of measurement points, and generate a Recipe Director-readable HSS File to realize automated measurement process.

Benefits of technology

The traditional 2-3-day measurement Recipe construction time is reduced to within 30 minutes, which improves product R&D efficiency and reduces manual intervention and machine occupancy time.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118501504B_ABST
    Figure CN118501504B_ABST
Patent Text Reader

Abstract

The present invention discloses a CD-SEM measurement method and device for semiconductor device layout graphics. The present invention obtains a test layout graphic file and at least one layer to be processed; de-layers the multiple layers to be processed; calculates geometric features according to polygon information in the measurement window; obtains the measurement type and measurement parameters of the measurement point position according to the geometric features; and compiles the measurement type, measurement parameters, positioning position, focus position, and measurement information parameters of the CD-SEM machine according to the measurement point into an HSS file format, and imports it into the CD-SEM machine for measurement. The present invention merges multiple layers into one layer and de-layers them, which facilitates multi-level analysis and takes into account the influence of the previous layer on the target point measurement method. The present invention does not require manual intervention, objectively analyzes the measurement type, positioning position, and focus position, liberates manual labor, and provides a feasible solution for the measurement of a large amount of CD-SEM graphic data.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of photolithography and manufacturing of semiconductor devices, and in particular relates to a CD-SEM (automated critical dimension scanning electron microscope) measurement method and equipment for semiconductor layout graphics. Background Art

[0002] In the field of integrated circuits, with the micro-machining and complexity of semiconductor device manufacturing processes, the use of scanning electron microscopes (SEM) for measurement and inspection is becoming increasingly important for semiconductor manufacturing. After each process step in the production of semiconductor manufacturing production lines, CD-SEM should be used to sample wafer products in a timely manner to determine whether the integrated circuits obtained after the steps of photolithography, exposure, development, etching, grinding, etc. meet the requirements, and then promote the subsequent steps. If abnormalities are found in the early stage, losses can be stopped in time. In the process of process research and development in the semiconductor field, the computational lithography and yield improvement departments will also frequently use CD-SEM to collect data. The computational lithography department collects a large number of CD-SEM images of test patterns to establish lithography models, etching models, model verification, etc., to ensure that there are no large deviations in the process parameters on the production line and to develop more advanced technology nodes in the future. These SEM images provide strong data support for lithography simulation verification. The yield improvement department also collects a large number of SEM images of hot spots and proposes targeted yield improvement methods.

[0003] CD-SEM images play an important role in both actual production and R&D. In order to collect a large number of SEM images, the traditional method is for engineers to find the measurement position on the wafer point by point on the CD-SEM machine, and then select the measurement method of the measurement position based on the engineer's experience, and select the appropriate positioning position and focus position according to the surrounding environment of the measurement position. The traditional method not only takes up precious CD-SEM machine time, but also wears out people's patience. An experienced measurement engineer takes about 1 minute to measure a single point, while the amount of CD-SEM image data required for an OPC model is thousands of valid measurement data, which will take two or three normal working days in total. Hitachi has developed a Recipe Director software to match its CD-SEM machine. The software can communicate with the CD-SEM. The software manually creates a measurement recipe based on the designed layout, and then imports it into the machine for measurement. This solves the problem of CD-SEM machine time constraints to a certain extent, and engineers can also operate in the office instead of entering the production line. However, the software still requires manual creation of measurement recipes point by point, and does not fundamentally solve the problem of collecting large amounts of data.

[0004] When measuring a point on a CD-SEM, selecting the positioning point, focus point, measurement method and measurement parameters are three main steps. Since the lens needs to move continuously between different measurement points during the CD-SEM measurement process, and any imaging of the lens will cause irreversible damage to the photoresist within the field of view, the positioning point, focus point and measurement point are three non-overlapping fields of view. Generally speaking, the positioning point and the focus point are the relative offset positions of the measurement point. The distance between the positioning point and the measurement point is no more than 6um and has relatively unique geometric characteristics compared with the surrounding environment. The focus point and the measurement point should be as close as possible but the field of view should not intersect. The field of view of the focus position should have both patterns and non-patterns. Therefore, based on the above rules, when measuring each point, the positioning point and the focus point, as well as the measurement method and parameters, need to rely on the experience of engineers to select appropriate values. This method relies heavily on manual labor and seriously hinders the production and R&D cycle of semiconductor processes. Summary of the invention

[0005] The purpose of the present invention is to address the deficiencies of the prior art and propose a CD-SEM measurement method and device for semiconductor device layout graphics to solve the problems of the current industry's collection of CD-SEM images, which is labor-intensive, consumes CD-SEM machine time, and relies on engineer experience. The method can directly reduce the CD-SEM measurement recipe construction process that takes 2 to 3 days to less than 30 minutes, accelerating product development efficiency.

[0006] The objective of the present invention is achieved through the following technical solutions:

[0007] In a first aspect, the present invention provides a CD-SEM measurement method for a semiconductor device layout pattern, comprising:

[0008] S1: Establishing a measurement graphic layout file, wherein the measurement graphic layout file includes multiple types of test layout graphics and their corresponding geometric features;

[0009] Preferably, the multiple types of test layout patterns include various polygons, namely lines, head-to-head, N*N square holes, and N*N rectangular holes;

[0010] The geometric features include horizontal geometric features and vertical geometric features; different test layout patterns have different geometric features, including but not limited to: the length, period and key size of lines, the gap and period of head-to-head, the length, width and gap of square holes, etc.;

[0011] S2: Set the measurement information parameters of the CD-SEM machine;

[0012] Preferably, the measurement information parameters of the CD-SEM machine include at least the following parameters: magnification, maximum jog length of OPC correction, maximum layout file size, measurement resolution, layout mask field of view, and the remaining parameters are stored in the form of templates on the CD-SEM machine for direct call;

[0013] S3: Obtain a test layout graphic file and at least one layer to be processed; design at least one test layout graphic on the layer to be processed according to the design requirements; if the layer to be processed is multi-layer, merge all the layers to be processed into one layer, and finally flatten the layout of all the layers to be processed, de-layer the multiple layers, and then execute step S4; if the layer to be processed is a single layer, directly execute step S4;

[0014] Optionally, if the layer to be processed (Layer) comes from a layout after optical proximity correction (OPC), jog removal is required;

[0015] S4: Get the measurement window;

[0016] Preferably, taking the custom measurement point as the center, a measurement window is obtained according to the magnification and measurement resolution;

[0017] S5: Calculate the geometric features at the measurement point according to the polygon information inside the measurement window;

[0018] Preferably, step S5 specifically comprises:

[0019] S5-1 traverses the polygon in the measurement window and uses the ray method to determine whether the measurement point is inside the polygon or on the edge of the polygon; if so, directly execute step 5-2; if not, perform a logical inverse operation on the polygon in the measurement window and replace the original polygon, and then execute step S5-2;

[0020] S5-2 uses a ray method to measure the transverse geometric features, wherein the transverse geometric features include the transverse key dimension length, the gap length on both sides of the transverse dimension, the measurement height, and the change trend of the transverse key dimension; the gap length on both sides of the transverse dimension includes the right gap length and the left gap length, and more preferably, specifically:

[0021] S5-2-1 issues a horizontal scan line with the measuring point as the center. The length of the scan line is the length of the measuring window, and the scan line width is the minimum unit of the software processing limit. The scan line intersects with all polygons in the measuring window to obtain multiple line segments. Each group of line segments is grouped into three categories, namely left line segments, middle line segments, and right line segments. The horizontal geometric characteristic parameters of the measuring point are analyzed and calculated. The horizontal critical dimension (CD) length at the measuring point is calculated based on the middle line segment, the right space (Right Space) length at the measuring point is calculated based on the right line segment, and the left space (RightSpace) length at the measuring point is calculated based on the left line segment.

[0022] S5-2-2 Scan upward and downward from the measuring point respectively, the scanning line length is the lateral key dimension, and the scanning line width is the minimum unit of the software processing limit, until the lateral key dimension change value obtained by scanning is greater than or equal to the threshold a1, calculate the upward scanning step length and the downward scanning step length, the total step length of the upward scanning step length and the downward scanning step length is the lateral CD corresponding measurement height, and then calculate the change trend of the lateral key dimension; for example, the threshold a1 is the OPC minimum correction jog length;

[0023] S5-3 uses a ray method to measure longitudinal geometric features, wherein the longitudinal geometric features include longitudinal key dimension length, longitudinal dimension gap lengths on both sides, measurement height, and longitudinal key dimension change trend; longitudinal dimension gap lengths on both sides include upper gap length and lower gap length; more preferably:

[0024] S5-3-1 Similarly, for the longitudinal features of the layout, repeat the above steps, issue a longitudinal scan line with the measurement point as the center, the scan line length is the measurement window length, and the scan line width is the minimum unit of the software processing limit; the scan line intersects with all polygons in the measurement window to obtain multiple line segments, and each group of line segments is grouped into 3 categories, namely upper line segments, middle line segments, and lower line segments, and analyze and calculate the longitudinal geometric feature parameters of the measurement point; calculate the longitudinal key dimension length at the measurement point based on the middle line segment, calculate the upper space length at the measurement point based on the upper line segment, and calculate the lower space length at the measurement point based on the lower line segment;

[0025] S5-3-2 Scan leftward and rightward from the measuring point respectively, the scanning line length is the longitudinal design key dimension, the scanning line width is the minimum unit of the software processing limit, until the longitudinal key dimension change value obtained by scanning is greater than or equal to the threshold a2, calculate the left scanning step length and the right scanning step length, the total step length of the left scanning step length and the right scanning step length is the longitudinal CD corresponding measurement height, and then calculate the change trend of the longitudinal key scale; for example, the threshold a2 is the OPC minimum correction jog length;

[0026] S6: Obtaining the measurement type and measurement parameters of the measurement point position according to the geometric characteristics of the measurement point;

[0027] Preferably, the measurement types include circle measurement, ellipse measurement, and two rectangle measurement;

[0028] First, determine whether the location of the measurement point is a rectangular polygon. If not, the measurement type adopts two-rectangle measurement. If so, continue to determine whether the horizontal key dimension is equal to the vertical key dimension. If they are equal, the measurement type adopts circle measurement, and the radius of the circle in the measurement parameters is set to half the length of the horizontal key dimension or the vertical key dimension. If they are not equal, then determine whether the ratio of the longer key dimension to the shorter key dimension is less than a threshold. If it is less than the threshold, the measurement type adopts ellipse measurement, and the major and minor axes of the ellipse in the measurement parameters are respectively set to the longer key dimension and the shorter key dimension. If it is greater than or equal to the threshold, the measurement type adopts two-rectangle measurement; wherein when the horizontal key dimension is not equal to the vertical key dimension, the longer dimension is selected from the horizontal key dimension and the vertical key dimension as the longer key dimension, and the other is the shorter key dimension;

[0029] If the measurement type adopts two-rectangle measurement, if the horizontal key dimension and the vertical key dimension are equal, the horizontal key dimension is selected as the key dimension Q; if the horizontal key dimension and the vertical key dimension are not equal, the shorter key dimension is selected as the key dimension Q; the key dimension Q is selected as the measurement direction, and it is determined whether the change trend of the key dimension Q is a decreasing trend. If it is a decreasing trend, the measurement type adopts the maximum value two-rectangle measurement, and the target key dimension and the measurement direction in the measurement parameters are the key dimension Q. The measurement length is the scanning line step length of the dimension corresponding to the key dimension Q, and the outward search length of the measurement frame is half of the length of the gap on both sides of the dimension corresponding to the key dimension Q; if it is an increasing trend, the measurement type adopts the minimum value two-rectangle measurement, and the mean value two-rectangle measurement is used in other cases.

[0030] S7: Obtaining the location of the measuring point (Adressing);

[0031] Preferably, taking the measuring point as the center, an observation window including the measuring window is intercepted;

[0032] Set a positioning window, divide the entire observation window into multiple (i.e., n*n) sub-windows, the sub-window size is the required positioning window size, traverse the observation window, and calculate the Fourier description operator of each sub-window;

[0033] The sub-window with the largest deviation from the mean in the Fourier description operator is regarded as the most outlier data point and used as the positioning window; the center of the positioning window is the positioning position;

[0034] S8: Get the focus position of the measurement point (Auto focus);

[0035] Preferably, in the observation window, the field of view around the measurement window is selected as the selected focus field of view, and the area ratio of the polygon in the selected focus field of view is calculated (that is, the area ratio of the polygon and the entire selected focus field of view in which it is located). If the ratio falls within the threshold range, the field of view is selected as the focus field of view, otherwise, the next selected focus field of view is found and the above operation is repeated;

[0036] S9: According to the measurement type, measurement parameters, positioning position, focus position, and measurement information parameters of the CD-SEM machine, the pandas open source library is used to organize the measurement points into an HSS file format that can be read by Recipe Director (measurement recipe creation software); the generated HSS (High-Speed ​​Streaming File, CD-SEM machine transmission format file) file is imported into Recipe Director, and further imported into the CD-SEM machine for actual measurement;

[0037] Preferably, if the measurement point is inside or on the edge of the polygon, the measurement point is corrected before step S6 until the two ends are at the same distance from the end of the polygon; if the measurement point is outside the polygon, the measurement point is corrected before step S6 until the two ends are at the same distance from the nearest polygon end.

[0038] In a second aspect, the present invention provides an electronic device, including a memory and a processor, wherein the memory stores executable code, and when the processor executes the executable code, the described method is implemented.

[0039] The beneficial effects of the present invention are as follows:

[0040] The present invention de-layers multiple layers by merging multiple layers into one layer and flattening the layout, so as to facilitate multi-layer analysis and take the influence of the previous layer on the target point measurement method into consideration.

[0041] The present invention extracts the geometric features of the measured position through the ray method and uses Fourier description operator window analysis to quickly and accurately screen out the optimal positioning position and focusing position.

[0042] The present invention can process multiple measurement points at one time, obtain the required measurement type, measurement parameters, positioning position, focus position, and measurement information parameter information of the CD-SEM machine, and use the pandas open source library to organize it into an HSS file format that can be read by Recipe Director; import the generated HSS file into Recipe Director, and further import it into the CD-SEM machine for actual measurement. The entire process does not require manual intervention, objectively analyzes the measurement type, positioning position, and focus position, liberates manual labor, and provides a feasible solution for the measurement of a large amount of CD-SEM graphic data. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 The figure is a flow chart of the method of the present invention.

[0044] Figure 2 Figure 1 is a test layout pattern, where (a)-(d) are common computational lithography test patterns in the test layout pattern, (a) is a line shape, (b) is a head-to-head shape, (c) is a hole or column, and (d) is a contact window hole; (e)-(h) are circuit layout patterns with actual circuit functions in the test layout pattern, all of which are randomly intercepted from the chip layout independently developed by Zhejiang University; (i)-(l) are TPGen randomly generated layout patterns in the test layout pattern, all of which are layout patterns randomly generated by the software kernel that do not violate the mask manufacturing rules (MRC) and have no practical significance.

[0045] Figure 3 A simplified diagram of the ray algorithm for determining the relative position of the measurement point, where (a) is the measurement point inside the polygon and (b) is the measurement point outside the polygon.

[0046] Figure 4 is the measurement window, where (a) is the measurement window and (b) is the logical exclusive OR of the measurement window.

[0047] Figure 5 The lateral geometric features of the measurement position are obtained by using the ray method, where (a) is a single lateral scan and (b) is a top-to-bottom frame-by-frame scan.

[0048] Figure 6 Different measurement types and measurement parameters, where (a) is circle measurement, (b) is ellipse measurement, (c) is minimum double rectangle measurement, (d) is maximum double rectangle measurement, and (e) is mean double rectangle measurement.

[0049] Figure 7 The best positioning position and the most suitable focusing position of the measurement point are shown in Figure 1, where (a) is the positioning field of view selected based on the Fourier description operator, and (b) is the focusing field of view selected based on the area ratio.

[0050] Figure 8 This is a schematic diagram of the sliding field of view window for selecting the best focus field of view position in this embodiment, where the numbers are the priorities of the fields to be selected, with the number 1 being the highest priority, the middle red five-pointed star position being the measurement target position, the red window being the measurement window, the green window being the focus field of view to be selected, and the field of view where the blue dot is located being the focus field of view finally selected.

[0051] Fig. 9 Create a metrology analysis example for one of the target locations in the layout.

[0052] Fig.10 It is a GUI interface software system corresponding to the method of the present invention. DETAILED DESCRIPTION

[0053] The specific implementation of the present invention will be described in more detail below in conjunction with the schematic diagram. The advantages and features of the present invention will become clearer based on the following description. It should be noted that the drawings are all in a very simplified form and are not in exact proportions, and are only used to facilitate and clearly assist in explaining the purpose of the embodiments of the present invention.

[0054] CD-SEM graphic measurement plays an important role in the fields of computational lithography and yield improvement. The computational lithography department collects a large number of CD-SEM images of test graphics to establish lithography models, etching models, model verification, etc. The yield improvement department also collects a large number of SEM images of hot spots for analysis. However, the best CD-SEM measurement solutions in the industry currently require engineers to analyze point by point in Recipe Director or directly on the CD-SEM machine to select the best measurement solution for each measurement data, including: finding points, selecting measurement conditions, selecting measurement methods, selecting positioning points, selecting focus points, etc. This traditional way of collecting CD-SEM measurement data is not conducive to lithography modeling data collection, nor is it conducive to random inspections of actual production. In order to solve the above problems, the present invention proposes a CD-SEM measurement method and equipment for semiconductor device layout graphics.

[0055] See also Figure 1 This embodiment provides a CD-SEM measurement method for a semiconductor device layout pattern, comprising the following steps:

[0056] Step S1: Provide the test pattern category for current lithography process verification, provide test pattern geometric feature parameters, set the pattern category parameter range, generate various predefined test pattern layouts, the geometric feature parameters of the test pattern layout will cover the input range, and then establish a measurement pattern layout file. Figure 2 In this case, the test pattern includes many types of test pattern patterns commonly used in computational lithography, see Figure 2Figures (a)-(d) in the middle: lines, head-to-head, N*N square holes, N*N rectangular holes and other polygons and circuit layout graphics with actual circuit functions, see Figure 2 Figures (e)-(h) in the middle, and the layout graphics randomly generated using Siemens EDA software TPGen, see Figure 2 Figures (i)-(l) in the middle. The graphics shown in the figures are all captured with the measurement window centered at the measurement position and a magnification of 100K. The graphic design unit is 1000dbu, and the graphic size is within 1um*1um.

[0057] The geometric features are inputs when designing test patterns. Different patterns have different geometric features, including but not limited to: the length, period and key dimensions of lines, the head-to-head gaps and periods, the length, width and gap of square holes, etc.

[0058] Step S2: setting the measurement information parameters of the CD-SEM machine; the CD-SEM measurement process requires setting a large number of related parameters. This method combines the layout feature analysis and only requires setting some of the parameters, including but not limited to: measurement magnification, focus magnification, positioning magnification, OPC corrected maximum jog unit length, maximum layout size, measurement resolution, layout mask field of view, etc. The remaining parameters will be stored in the form of templates on the CD-SEM machine and can be directly called.

[0059] Step S3: Obtain the test layout graphic file and at least one layer to be processed (Layer), and use the Klayout open source layout processing library to obtain them; design at least one test layout graphic on the layer to be processed according to the design requirements; if the layer to be processed (Layer) is multi-layer, merge all the layers to be processed (Layer) into one layer, analyze the mutual characteristics between the multiple layers to obtain more accurate measurement methods and measurement parameters, and finally flatten the layout of all the layers to be processed (Layer), de-layer the multiple layers, and then execute step S4; otherwise, if the layer to be processed (Layer) is a single layer, directly execute step S4.

[0060] Optionally, if the layer to be processed (Layer) comes from the layout after OPC correction, jog (small corner) removal is required;

[0061] Step S4: Get the measurement window:

[0062] Centered on the custom measurement point, after preprocessing in step S3, the layout uses klayout to obtain the measurement window according to the magnification and measurement resolution. The window contains the measurement location information and surrounding environment information.

[0063] The measurement window size is calculated as follows:

[0064] Formula (1)

[0065] Mag is the magnification, Res is the image resolution; a 6um window is captured to facilitate calculation of the best positioning and focusing positions. It is recommended that the CD-SEM machine positioning should not exceed 6um beyond the measurement position.

[0066] Step S5: Calculate the geometric features at the measurement point according to the polygon information inside the measurement window; the geometric features include transverse geometric features and longitudinal geometric features; specifically:

[0067] Provide the above-mentioned measurement window and its internal polygon information, and perform geometric feature analysis on the measurement position. The horizontal geometric features include the horizontal key dimension (Horizontal Design CD), right space (Right Space), left space (Right Space), upper length (Up Length), lower length (Down Length), the horizontal key dimension corresponding to the measurement height (Up Length + Down Length), and the horizontal key dimension change trend. The vertical geometric features include the vertical key dimension (Vertical Design CD), upper space (Up Space), lower space (Down Space), left length (LeftLength), right length (Right Length), the vertical key dimension corresponding to the measurement width (Left Length + RightLength), and the vertical key dimension change trend. The specific implementation method is:

[0068] i) See Figure 3 , the number in the figure indicates the nth intersection point between the ray and the polygon. Traverse the polygon in the measurement window and use the ray method to determine whether the measurement point is inside the polygon (it is also considered inside on the edge of the polygon). Draw a horizontal ray from the measurement point and count the number of intersections between the ray and the polygon. If the number of intersections is an odd number, see Figure 3 (a) in the figure, the point is inside the polygon or on the edge, and step ii) is performed. If the number of intersection points is even, see Figure 3 In Figure (b), the point is outside the polygon.

[0069] Optionally, see Figure 4 In (a)-(b), if the measurement point is outside the polygon, the polygon in the measurement window is logically inverted and replaced with the original polygon in the window, and then step ii is executed. The specific implementation method is: a rectangle of the same size as the measurement window is designed, and the rectangle and the polygon in the window are XORed at the same time, and the output result is the graphic logical inversion result.

[0070] ii) Using the ray method to obtain the geometric characteristics of the measurement position, the specific implementation method is shown in Figure 5 .

[0071] ① Horizontal geometric characteristics

[0072] A horizontal scan line is sent out with the measurement point as the center. The scan line length is the measurement window length, and the scan line width is the minimum unit of the software processing limit. The scan line intersects with the polygon in the measurement window to obtain multiple line segments. Figure 5 In the figure (a), each group of line segments is divided into three categories. There are three line segments on the left, namely A1, A2, and A3. The middle line segment is the only line segment C1 that intersects the measurement position. There are two line segments on the right, namely B1 and B2.

[0073] The lateral critical dimension at the measuring point is calculated based on the middle line segment (C1), which is the length of the middle line segment.

[0074] Calculate the difference between the minimum value of the horizontal coordinate of the first line segment on the right (B1) and the maximum value of the horizontal coordinate of the middle line segment (C1). The result is the right space.

[0075] Right Space = min(x(B1)) – max(x(C1)) Formula (2)

[0076] Calculate the difference between the minimum horizontal coordinate value of the middle line segment (C1) and the maximum horizontal coordinate value of the first line segment on the left (A1). The result is the left space.

[0077] Left Space = min(x(C1)) – max(x(B1)) Formula (3)

[0078] See also Figure 5 In the middle (b) figure, the scan is performed upward and downward from the measurement position, and the scan line length is the lateral critical dimension, and the scan line width is the minimum unit of the software processing limit, until the lateral CD change obtained by scanning is greater than the threshold a1. In this case, the threshold a1 is set to the maximum movement unit of the OPC correction, and the upward or downward movement length of the scan line is calculated, that is, the upward scanning step length and the downward scanning step length at this point, respectively. The total step length of the above two scanning steps is the corresponding measurement height of the lateral CD. The AC line segment in the figure is the upward scanning length. The corresponding relationship between the CD after the change and the lateral critical dimension initially calculated is recorded to summarize the change trend of the critical dimension. In the figure, the critical dimension becomes smaller after scanning upward.

[0079] ② Longitudinal geometric characteristics

[0080] By repeating the above step ①, the longitudinal key dimensions, upper and lower gaps, left and right lengths, corresponding measured width and CD change trend are calculated.

[0081] Similarly, for the longitudinal features of the layout, repeat the above steps, issue a longitudinal scan line with the measurement point as the center, the scan line length is the measurement window length, and the scan line width is the minimum unit of the software processing limit; the scan line intersects with all polygons in the measurement window to obtain multiple line segments, and each group of line segments is grouped into 3 categories, namely upper line segments, middle line segments, and lower line segments, and analyze and calculate the longitudinal geometric feature parameters of the measurement point; calculate the longitudinal key dimension length at the measurement point based on the middle line segment, calculate the upper space length at the measurement point based on the upper line segment, and calculate the lower space length at the measurement point based on the lower line segment;

[0082] Scan leftward and rightward from the measuring point, respectively. The scanning line length is the longitudinal design key dimension, and the scanning line width is the minimum unit of the software processing limit. When the longitudinal key dimension change value obtained by scanning is greater than or equal to the threshold a2, the left scanning step length and the right scanning step length are calculated. The total step length of the above two scanning steps is the longitudinal CD corresponding measurement height, and then the change trend of the longitudinal key scale is calculated; for example, the threshold a2 is the minimum correction jog length of OPC;

[0083] Optionally, the measurement point can be corrected before step S6: if the measurement point is inside or on the edge of the polygon, the measurement point is corrected until both ends are at the same distance from the end of the polygon; if the measurement point is outside the polygon, the measurement point is corrected until both ends are at the same distance from the nearest polygon end.

[0084] S6: Obtaining the measurement type and measurement parameters of the measurement point position according to the geometric characteristics of the measurement point;

[0085] The measurement types include circle measurement, ellipse measurement, and two-rectangle measurement; the two-rectangle measurement includes minimum value double-rectangle measurement, maximum value double-rectangle measurement, and mean value double-rectangle measurement.

[0086] First determine whether the location of the measurement point is a rectangular polygon. If not, the measurement type uses two-rectangle measurement. Figure 6 In Figures (c)-(e), if yes, then continue to determine whether the horizontal key dimension is equal to the vertical key dimension. If yes, the measurement type is circular measurement, see Figure 6 In the middle (a) figure, the radius of the circle in the measurement parameters is set to half the length of the horizontal key dimension or the vertical key dimension. If they are not equal, it is determined whether the ratio of the longer key dimension to the shorter key dimension is less than the threshold (set to 1.2 in this case). If it is less than the threshold, the measurement type is elliptical measurement, see Figure 6In the middle (b) figure, the major and minor axes of the ellipse in the measurement parameters are set as the longer key dimension and the shorter key dimension respectively. If they are greater than or equal to the threshold, the measurement type adopts two-rectangle measurement; when the horizontal key dimension is not equal to the vertical key dimension, the longer dimension is selected from the horizontal key dimension and the vertical key dimension as the longer key dimension, and the other is the shorter key dimension.

[0087] If the measurement type adopts two-rectangle measurement, if the horizontal key dimension and the vertical key dimension are equal, the horizontal key dimension is selected as the key dimension Q; if the horizontal key dimension and the vertical key dimension are not equal, the shorter key dimension is selected as the key dimension Q. The key dimension Q is selected as the measurement direction to determine whether the change trend of the key dimension Q is a decreasing trend. If it is a decreasing trend, the measurement type adopts the maximum value two-rectangle measurement, refer to Figure 6 In the middle (c) figure, the target key dimension and measurement direction in the measurement parameters are the key dimension Q and its corresponding dimension, the measurement length is the measurement height of the dimension corresponding to the key dimension Q, and the outward search length of the measurement frame is half of the length of the gap on both sides of the dimension corresponding to the key dimension Q; if it is a larger trend, the measurement type adopts the minimum value two rectangle measurement, see Figure 6 Figure (d). Figure 6 In the middle (e) figure, the mean two-rectangle measurement is used in other cases.

[0088] It is worth noting that in actual situations, determining the measurement method and various measurement parameters is more complicated than what is described above. The present invention adopts a very simplified form and uses non-precise proportions, which are only used to conveniently and clearly assist in explaining the purpose of the embodiments of the present invention.

[0089] S7: Get the location of the measuring point:

[0090] Taking the measurement point as the center, cut out an observation window including the measurement window;

[0091] Set a 6um positioning window, divide the entire observation window into n sub-windows, the sub-window size is the window size required by Adressing, in this embodiment it is 1um*1um, and the division step is 100nm*100nm. Traverse the observation window and calculate the Fourier description operator of each sub-window;

[0092] The Fourier description operator is as follows:

[0093] Formula (4)

[0094] Where K is the number of description operators, which is 18 in this embodiment, k is the number of polygon boundary points, s(k)=x(k)+jy(k); j represents an imaginary number; x(k) represents the horizontal coordinate of the kth point, y(k) is the vertical coordinate, and u is the Fourier series, u=1, 2,··, +∞;

[0095] The sub-window with the largest deviation from the mean in the Fourier description operator is regarded as the most outlier data point and used as the positioning window; the center of the positioning window is the positioning position;

[0096] It is worth noting that the sub-windows that overlap with the measurement window should be excluded during calculation to avoid overlapping of the field of view and damage to the photoresist at the measurement position. Figure 7 In the middle figure (a), the algorithm selects a nearby cross as the positioning position.

[0097] S8: Get the focus position of the measurement point:

[0098] Provide an observation window of 6um size. In the observation window, slide and select the field of view around the measurement window as the candidate focus field of view. During the sliding process, follow the principle of closest to the measurement field of view to traverse, where the number is the selection priority, 1 is the first candidate field of view, and so on. The focus field of view implemented in this case is 500nm*500nm. Calculate the ratio of the polygon area in the field of view to the area of ​​the candidate focus field of view. If it falls within the threshold range, the field of view is selected as the focus field of view. If the ratio is between 20% and 80%, refer to Figure 7 (b) in the middle, otherwise it will look for the next candidate focus field; see Figure 8 , the candidate field of view 193 closest to the measurement position is selected, and the field of view with a graphic area ratio of 30% is used as the focus position, while the graphic area ratio of field of view 46 and field of view 47 is too small and is not selected as the final choice, even though they are closer to the measurement window.

[0099] S9: According to the measurement type, measurement parameters, positioning position, focus position, and measurement information parameters of the CD-SEM machine, the pandas open source library is used to organize them into an HSS file format that can be read by Recipe Director; the present invention can perform multi-measurement point analysis and write HSS files at the same time to achieve multi-point measurement. The generated HSS file is imported into RecipeDirector and further imported into the CD-SEM machine for actual measurement.

[0100] See also Fig. 9, a head-to-head graphic measurement is required. According to the process of automatically establishing the CD-SEM measurement method of the present invention, a 6um*6um observation window is first captured, in which window A is the measurement window, window B is the focus window obtained according to the graphic density ratio and does not overlap with window A, and window C is the positioning window with the most special graphic features obtained by using the Fourier description operator.

[0101] An embodiment of the present invention provides an electronic device. Specifically, the electronic device includes a memory and a processor. The memory stores executable code. When the processor executes the executable code, the method described in any one of the implementation modes is implemented.

[0102] The memory may include high-speed random access memory (RAM), and may also include non-volatile memory (Non-volatile Memory), such as at least one disk storage. The communication connection between the system network element and at least one other network element is realized through at least one communication interface (which may be wired or wireless), and the Internet, wide area network, local area network, metropolitan area network, etc. may be used.

[0103] The bus may be an ISA bus, a PCI bus or an EISA bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc.

[0104] Among them, the memory is used to store the program, and the processor executes the program after receiving the execution instruction. The method executed by the device for flow process definition disclosed in any embodiment of the above-mentioned embodiment of the present invention can be applied to the processor or implemented by the processor.

[0105] The processor may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by an integrated logic circuit of hardware in the processor or an instruction in the form of software. The above processor may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of the present invention can be implemented or executed. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The steps of the method disclosed in conjunction with the embodiments of the present invention can be directly embodied as being executed by a hardware decoding processor, or executed by a combination of hardware and software modules in a decoding processor. The software module may be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory, and the processor reads the information in the memory and completes the steps of the above method in combination with its hardware.

[0106] Furthermore, the present invention also designs a software system for the above method based on the Windows operating system, see Fig.10, the software includes an easy-to-operate GUI interface, and the software includes the simplified operation of automatically establishing the CD-SEM measurement recipe, which is convenient for people who are not professional in the development environment to use the method. The interface software runs on the Window system and does not need to be installed. When the processor executes the executable code, the method and system for automatically building CD-SEM graphic measurement based on layout feature analysis are implemented. Among them, the Select a handbook control is to select a measurement coordinate text file, which describes the position and measurement point naming that the user wants to measure on the wafer. The GDS control is to select the layout file location corresponding to the wafer to be measured. The Reset control is used to initialize the entire software system. The Unknow drop-down box selects the names of different measurement parameter templates saved on the CD-SEM machine, including EPS templates, MP templates, alignment templates, etc. The Dark tone check box selects the bright field or dark field field of view of the wafer corresponding to the mask. Origin is the offset of the measurement point relative to the origin of the layout. Layer is the layer to be measured and analyzed, which can be multiple layers. When the above files and initialization are ready, click Generate to generate an HSS file that can be read by Recipe Director.

Claims

1. A CD-SEM measurement method for semiconductor device layout graphics, characterized in that The method comprises: Step S1: establishing a measurement graphic layout file, wherein the measurement graphic layout file includes multiple types of test layout graphics and their corresponding geometric features; Step S2: Setting the measurement information parameters of the CD-SEM machine; Step S3: Obtain a test layout graphic file and at least one layer to be processed; design at least one test layout graphic on the layer to be processed according to the design requirements; if the layer to be processed is multi-layer, merge all the layers to be processed into one layer, and finally flatten the layout of all the layers to be processed, de-layer the multiple layers, and then execute step S4; if the layer to be processed is a single layer, directly execute step S4; Step S4: obtaining a measurement window; Step S5: Calculate the geometric features at the measurement point according to the polygon information inside the measurement window; the geometric features include transverse geometric features and longitudinal geometric features; specifically: S5-1 traverses the polygon in the measurement window and uses the ray method to determine whether the measurement point is inside the polygon or on the edge of the polygon; if so, directly execute step 5-2; if not, perform a logical inverse operation on the polygon in the measurement window and replace the original polygon, and then execute step S5-2; S5-2 uses the ray method to measure the lateral geometric features, wherein the lateral geometric features include the lateral key dimension length, the right gap length, the left gap length, the measurement height, and the change trend of the lateral key dimension; specifically: S5-2-1 issues a horizontal scan line with the measuring point as the center. The scan line length is the measuring window length, and the scan line width is the minimum unit of the software processing limit. The scan line intersects with all polygons in the measuring window to obtain multiple line segments, and each group of line segments is grouped into three categories, namely left line segments, middle line segments, and right line segments. The horizontal key dimension length of the measuring point is calculated based on the middle line segment, the right gap length of the measuring point is calculated based on the right line segment, and the left gap length of the measuring point is calculated based on the left line segment. S5-2-2 Scan upward and downward from the measuring point respectively, the scanning line length is the length of the lateral key dimension, and the scanning line width is the minimum unit of the software processing limit, until the lateral key dimension change value obtained by scanning is greater than or equal to the threshold a1, calculate the upward scanning step length and the downward scanning step length, the total step length of the upward scanning step length and the downward scanning step length is the measuring height, and then calculate the change trend of the lateral key dimension; S5-3 uses the ray method to measure the longitudinal geometric features, wherein the longitudinal geometric features include the longitudinal key dimension length, the upper gap length, the lower gap length, the measurement height, and the change trend of the longitudinal key dimension; specifically: S5-3-1 issues a longitudinal scan line with the measuring point as the center, the scan line length is the measuring window length, and the scan line width is the minimum unit of the software processing limit; the scan line intersects with all polygons in the measuring window to obtain multiple line segments, and each group of line segments is grouped into three categories, namely, upper line segments, middle line segments, and lower line segments; the longitudinal key dimension length of the measuring point is calculated based on the middle line segment, the upper gap length of the measuring point is calculated based on the upper line segment, and the lower gap length of the measuring point is calculated based on the lower line segment; S5-3-2 Scan leftward and rightward from the measuring point respectively, the scanning line length is the longitudinal key dimension length, the scanning line width is the minimum unit of the software processing limit, until the longitudinal key dimension change value obtained by scanning is greater than or equal to the threshold a2, calculate the left scanning step length and the right scanning step length, the total step length of the left scanning step length and the right scanning step length is the measuring height, and then calculate the change trend of the longitudinal key dimension; Step S6: obtaining the measurement type and measurement parameters of the measurement point position according to the geometric features of the measurement point; Step S7: Obtain the location of the measurement point; specifically: Taking the measurement point as the center, cut out an observation window including the measurement window; Divide the entire observation window into multiple sub-windows, traverse the observation window, and calculate the Fourier description operator of each sub-window; The sub-window with the largest deviation from the mean in the Fourier description operator is regarded as the most outlier data point and used as the positioning window; the center of the positioning window is the positioning position; Step S8: Obtain the focus position of the measurement point; specifically: In the observation window, the field of view around the measurement window is selected as the candidate focus field of view, and the area ratio of the polygon area in the candidate focus field of view is calculated. If the ratio falls within the threshold range, the field of view is selected as the focus field of view, otherwise, the operation is repeated for the next candidate focus field of view; Step S9: Arrange the measurement points into an HSS file format readable by Recipe Director according to the measurement type, measurement parameters, positioning position, focus position, and measurement information parameters of the CD-SEM machine; import the generated HSS file into RecipeDirector, and then further import it into the CD-SEM machine for measurement.

2. The method according to claim 1, characterized in that: In step S2, the measurement information parameters set for the CD-SEM machine include at least the following parameters: magnification, maximum jog length of OPC correction, maximum layout file size, measurement resolution, and layout mask field of view.

3. The method according to claim 1, characterized in that: The measurement types in step S6 include circle measurement, ellipse measurement, and two rectangle measurement.

4. The method according to claim 1, characterized in that: If the measuring point is inside the polygon or on the edge of the polygon, the measuring point is corrected before step S6 until the two ends are at the same distance from the end of the polygon; if the measuring point is outside the polygon, the measuring point is corrected before step S6 until the two ends are at the same distance from the nearest polygon end.

5. An electronic device, characterized in that: The invention comprises a memory and a processor, wherein the memory stores executable codes, and when the processor executes the executable codes, the method according to any one of claims 1 to 4 is implemented.