Wafer positioning detection method and wafer positioning detection system

Abstract

The application provides a wafer positioning detection method and a wafer positioning detection system, and relates to the technical field of semiconductors.The method comprises driving a wafer to move on a machine table in a preset path;starting a scanning acquisition card to obtain reflected light of a laser beam irradiating a wafer surface; sampling wafer surface information in the form of scattered points on an Archimedes spiral from a preset starting point and caching the wafer surface information as a scanning image; the scanning image comprises an edge image of an edge range of the scanned wafer; when the wafer surface information of the edge image completely covers the edge range of the wafer, a calibration deviation value of the wafer is obtained according to the cached edge image, and a current position coordinate of the wafer is calibrated as a calibration position coordinate; the scanning image is converted and spliced after conversion, and a wafer scanning image under the calibration position coordinate is obtained, so that wafer alignment correction and wafer defect detection can be realized in the same scanning process, and wafer detection efficiency is significantly improved.

Description

Technical Field

[0001] This application relates to the field of semiconductor technology, and more specifically, to a wafer positioning and detection method and a wafer positioning and detection system. Background Technology

[0002] A wafer refers to a silicon chip used in the fabrication of integrated circuits (ICs). Various circuit element structures can be fabricated on a wafer, and further packaging processes can be performed to create integrated circuit products with specific electrical functions. In the IC industry, with the development of high-density, multi-functional integrated circuits, wafers need to integrate more and more devices and circuit element structures within their limited area, which places increasingly higher demands on the precise positioning during wafer fabrication.

[0003] Large-scale integrated circuit manufacturing and testing processes require the use of semiconductor equipment. After wafer mounting, most equipment requires wafer alignment (WA). Currently, the industry typically uses optical imaging systems to capture sample images of the wafer for centering and calibration. A robotic arm places the wafer in the working position, and firstly, a planar array camera, line array camera, or point sensor samples preset points on the wafer edge to obtain sample images. By processing the information in the sample images, the position information of the wafer edge points is obtained. Then, circle fitting and line fitting are performed to obtain alignment results, including the center position and orientation angle. After correcting the current position of the wafer based on the alignment results, a full scan of the wafer is performed to execute core inspection and measurement tasks.

[0004] This positioning and detection method suffers from inaccurate positioning and slowed work efficiency. On the one hand, due to differences in wafer size, setting position and orientation, and the limitation of the field of view of the optical imaging system, the pre-sampling points may not be able to sample a complete wafer edge image, resulting in inaccurate alignment results obtained through fitting calculation. On the other hand, measuring the edge position of the wafer first and fitting calculation to obtain the alignment result, then calibrating the wafer position based on the alignment result before wafer inspection takes a long time, affecting the work efficiency of wafer defect detection. Summary of the Invention

[0005] This application provides a wafer positioning and inspection method and a wafer positioning and inspection system, which can realize wafer alignment correction and wafer defect detection in the same scanning process, significantly improving wafer inspection efficiency.

[0006] To achieve the above objectives, the technical solutions adopted in the embodiments of this application are as follows: One aspect of this application provides a wafer positioning and detection method, comprising: driving a wafer to move on a machine along a preset path; activating a scanning acquisition card to acquire reflected light from a laser beam irradiating the wafer surface, sampling wafer surface information from a preset starting point in the form of scattered points on an Archimedean spiral and buffering it as a scanning image, the scanning image including an edge image of the edge range of the scanned wafer; when the wafer surface information of the edge image completely covers the edge range of the wafer, obtaining a calibration deviation value of the wafer based on the buffered edge image, calibrating the current position coordinates of the wafer to calibration position coordinates; converting the scanning image and stitching the converted scanning images together to obtain a wafer scanning image at the calibration position coordinates.

[0007] In one feasible implementation of the embodiments of this application, driving the wafer to move on the machine platform along a preset path includes: fixing the wafer on the machine platform; driving the machine platform to move linearly along a first direction and rotating around the geometric center of the wafer, wherein the first direction is any linear direction on the plane where the machine platform is located.

[0008] In one feasible implementation of this application, sampling wafer surface information in the form of scattered points on an Archimedean spiral from a preset starting point and caching it as a scanned image includes: sampling in the form of scattered points on an Archimedean spiral from a preset starting point to obtain a sampled image containing wafer surface information at the current scattered point position, and caching the sampled image; arranging the sampled images corresponding to the scattered points on each turn of the Archimedean spiral into a row, and multiple rows into columns to cache the sampled images as a two-dimensional Mercator diagram.

[0009] In one feasible implementation of this application, before starting the scanning acquisition card to acquire the reflected light irradiated by the laser beam on the wafer surface, the method further includes: determining the edge contour points of the current wafer based on the current wafer size information and pre-stored historical data information, and using the edge contour points as preset starting points.

[0010] In one feasible implementation of the embodiments of this application, when the wafer surface information of the edge image completely covers the edge range of the wafer, the wafer's outer edge contour includes a closed-loop adjacent round edge and a marked edge, the marked edge includes a flat-cut edge or a V-shaped cut edge, the wafer's edge range includes a round edge contour and a flat-cut edge contour, or the wafer's edge range includes a round edge contour and a V-shaped cut pattern.

[0011] In one feasible implementation of this application, obtaining the wafer calibration deviation value based on the cached edge image and calibrating the wafer's current position coordinates to calibration position coordinates includes: segmenting and extracting circular edge pixels and marker edge pixels in the edge image; obtaining the wafer calibration deviation value, wherein the wafer's center position is obtained based on the circular edge pixels, and the wafer's orientation angle is obtained based on the marker edge type and marker edge pixels; adjusting the wafer's coordinate position based on the calibration deviation value, and / or adjusting the wafer's coordinate system based on the calibration deviation value to calibrate the wafer's current position coordinates to calibration position coordinates.

[0012] In one feasible embodiment of this application, converting the scanned image and stitching the converted scanned image includes: reconstructing the Mercator image into a wafer image in Cartesian coordinates, wherein the wafer is mapped to an M-axis according to the wafer size and the pixel values ​​of the stitching. The image contains M pixels. The pixel values ​​of the sub-image in row i and column j are calculated by bilinear interpolation of the corresponding scanned image in the Mercator image. The position of the sub-image in row i and column j is obtained by translating and rotating the scanned image and then transforming it into Cartesian coordinates. The sub-images are then stitched together to obtain the wafer image.

[0013] In one feasible embodiment of this application, the wafer positioning detection method further includes: when the wafer surface information of the edge image completely covers the edge range of the wafer, the scanning acquisition card continues to acquire the reflected light irradiated by the laser beam on the wafer surface, and acquires the center image of the wafer surface in the form of scattered points on the Archimedean spiral.

[0014] Another aspect of this application provides a wafer positioning and detection system, including a wafer carrier stage, a motion mechanism mounted on the wafer carrier stage, a laser source, a receiving sensor, and a processor. The motion mechanism includes a horizontal motion platform that drives the wafer carrier stage to move in a planar manner and a rotation axis that drives the wafer carrier stage to rotate horizontally. The laser source emits a laser beam toward the wafer mounted on the wafer carrier stage, so the receiving sensor receives the laser beam reflected by the wafer to form a scan image of the current detection position on the wafer. The processor is electrically connected to the motion mechanism, the laser source, and the receiving sensor, and is used to control the operation of the laser source and the motion mechanism, and to buffer, convert, and stitch the scan image received by the receiving sensor.

[0015] In one feasible implementation of this application, the processor extracts the edge image covering the edge range of the wafer to obtain the wafer calibration deviation value, calculates the wafer calibration coordinate position based on the calibration deviation value, and transforms and stitches the scanned image based on the wafer calibration coordinate position to obtain the wafer scanned image under the calibration position coordinates.

[0016] This application provides a wafer positioning and detection method and a positioning and detection system. The wafer positioning and detection method includes driving the wafer to move on a machine along a preset path; activating a scanning acquisition card to acquire reflected light from a laser beam irradiating the wafer surface; sampling wafer surface information from a preset starting point in a scattered manner along an Archimedean spiral and buffering it as a scanning image; the scanning image includes an edge image of the edge range of the scanned wafer; when the wafer surface information of the edge image completely covers the edge range of the wafer, obtaining the wafer calibration deviation value based on the buffered edge image, and calibrating the current position coordinates of the wafer to calibration position coordinates; converting the scanning image and stitching the converted scanning images together to obtain a wafer scanning image at the calibration position coordinates. When the wafer surface information of the edge image completely covers the edge range of the wafer, the cached edge image can be calculated and processed simultaneously during the scanning and caching process to obtain the wafer calibration deviation value. Based on the calibration deviation value, the current position coordinates of the wafer are calibrated to the calibration position coordinates. The cached scan image is then converted into a scan image under the calibration position coordinates, and the converted scan images are stitched together to obtain the wafer scan image under the calibration position coordinates. This enables wafer alignment correction and wafer defect detection during the sampling and caching process of the scan image. While completing the scan image, the wafer position calibration is completed and the output is a wafer scan image under the calibrated coordinate position, which significantly improves the efficiency of wafer alignment and wafer inspection. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic diagram of the structure of a semiconductor device in the prior art; Figure 2 This is a schematic diagram of wafer placement position deviation in a wafer positioning detection method provided in an embodiment of this application; Figure 3a A wafer positioning and detection method provided in this application includes a method for setting wafer alignment sampling points; Figure 3b Another method for setting wafer alignment sampling points in a wafer positioning and detection method provided in this application embodiment; Figure 4 This is one of the flowcharts for a wafer positioning and detection method provided in an embodiment of this application; Figure 5This is a schematic diagram of the path of the Archimedean spiral in a wafer positioning and detection method provided in an embodiment of this application; Figure 6 A second flowchart illustrating a wafer positioning and detection method provided in this application embodiment; Figure 7a This is a schematic diagram illustrating the edge range of the wafer covered by a scanned image in a wafer positioning and detection method provided in an embodiment of this application. Figure 7b This is a schematic diagram illustrating the coverage of the center area of ​​the wafer by a scanned image in a wafer positioning and detection method provided in an embodiment of this application. Figure 7c This is a schematic diagram illustrating a wafer positioning and detection method provided in an embodiment of this application, where the scanned image covers the entire wafer. Figure 8a This is a schematic diagram of the outer contour of a wafer in a wafer positioning and detection method provided in an embodiment of this application; Figure 8b This is another schematic diagram of the outer contour of a wafer in a wafer positioning and detection method provided in an embodiment of this application; Figure 9 This is a partial schematic diagram of the edge image of a wafer in a wafer positioning and detection method provided in an embodiment of this application; Figure 10 This is a schematic diagram of a wafer positioning and detection system provided in an embodiment of this application.

[0019] Icons: 100 - Semiconductor equipment; 110 - Front-end mechanical module; 112 - Wafer cassette; 113 - Wafer pre-aligner; 114 - Robotic arm; 115 - Mechanical motion platform; 120 - Wafer; 130 - Optical microscopic imaging system; 140 - Computer; 141 - Software; 201 - Predetermined position of the wafer; 202 - Actual placement position of the wafer; 203 - Positional deviation of wafer placement; 204 - Wafer orientation angle deviation; 301, 302, 303 - Sampling points; 311, 312, 313 - Sampling point at the arc edge; 314, 315 - Sampling point at the flat edge; 401 - Receiver sensor; 402 - Wafer carrier stage; 403 - Motion mechanism; 404 - Wafer under test; 405 - Current sampling point; 406 - Laser source; 501 - Preset starting point; 502 - Spacing between two adjacent turns in the Archimedean spiral; 503 - End point; 701 - Edge range; 702 - Center range; 703 - Wafer range; 801 - Arc edge sampling sub-image; 802 - Flat edge sampling sub-image; 803 - Arc edge sampling sub-image. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. It should be noted that, in the absence of conflict, the various features in the embodiments of this application can be combined with each other, and the combined embodiments are still within the protection scope of this application.

[0021] In the description of this application, it should be noted that the terms "upper," "lower," "left," "right," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing this application and simplifying the description, and therefore should not be construed as limiting this application. In addition, the terms "first," "second," and "third," etc., are only used to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0022] like Figure 1 As shown, a conventional semiconductor device 100 typically includes a front-end mechanical module 110 for wafer loading and unloading, which can hold a wafer cassette 112. This module also includes a wafer pre-aligner 113 and a robotic arm 114, which manipulates the loading and unloading of wafers. It also includes a mechanical motion platform 115, which is typically movable in the X, Y, and Z directions and rotates around the Z direction. A wafer tray (e-chuck) can be placed on the mechanical motion platform 115 for more stable wafer fixation 120. The semiconductor device 100 also includes a computer 140 and software 141 running thereon, including user interface (GUI), system software and algorithm software, hardware control, and communication modules. It also includes an optical microscope (OM) 130 for positioning and detection, used for wafer alignment, wafer defect image acquisition, linewidth measurement, and overlay measurement.

[0023] During operation, the wafer 120 is first pre-aligned by the front-end mechanical module 110 and then mounted onto the mechanical motion platform 115. Due to the mechanical motion error of the robot arm 114 and the mechanical motion platform 115, there will be deviations between the wafer center position and the pre-determined contour position after each mounting.

[0024] like Figure 2 As shown, the positional deviation 203 (dx, dy) between the actual placement position 202 and the predetermined position 201 of the wafer is typically within the range of 200 μm, and the wafer orientation angle deviation 204 between the actual placement position 202 and the predetermined position 201 is typically within the range of ±1°. Such a degree of deviation is obviously an unacceptable error range in the process of integrated circuit manufacturing. Therefore, wafer calibration and alignment must be performed in advance to ensure the working accuracy when performing subsequent detection or measurement tasks.

[0025] Existing wafer alignment technologies, such as Figure 3a As shown, sampling points are pre-defined in the edge region of the wafer. The optical imaging system uses an area array camera to sample and capture images of these sampling points. Sampling points 301 and 302 are both arc edges of the wafer. After the camera captures the images, the data is processed in a computer to obtain the edge location points. A least-squares calculation of the circle is then performed on these edge location points to obtain the center and radius of the wafer. Sampling point 303 is a sampling point for the flat edge notch of the wafer. After the camera captures the images, a straight line fit is performed on the edge to obtain the wafer's orientation angle. Another alignment method is as follows... Figure 3b As shown, in this method, the optical microscopic imaging system uses a linear array camera or point sensor to capture images of the wafer edge. Points 311, 312, and 313 are sampling points for the arc edge, and points 314 and 315 are sampling points for the flat edge. After sampling, the image is processed to obtain the position of the wafer edge points, and then circle fitting and line fitting are performed to obtain the alignment result: the center position and orientation angle. However, both of the above alignment methods have problems and defects. For example... Figure 3a As shown, the alignment method requires pre-dividing sampling points according to the wafer size. This pre-division is based on ideal positioning. However, since the actual wafer placement and orientation may differ, and the field of view of optical imaging systems is typically small, it may result in incomplete wafer edge images being obtained through the sampling points. This necessitates adjusting the wafer position using a mechanical motion platform and then supplementing the sampling images. This rework significantly reduces inspection efficiency. Furthermore, combined with... Figure 3b As shown, even without considering rework due to incomplete sampling, the alignment process typically involves separately capturing images of the curved edges and notched edges of the wafer, processing the images to obtain the edge point positions, and then fitting a circle or straight line to obtain the specific alignment result. The calculation process for fitting a circle or straight line requires a large amount of edge point data to improve accuracy and avoid noise points from being included in the calculation, which places high demands on the computer's computing power. Furthermore, the excessive time consumed during wafer testing significantly slows down the efficiency of wafer positioning and inspection.

[0026] To address the aforementioned problems, this application provides a wafer positioning and detection method, such as... Figure 4 As shown, the method includes: S101, drive the wafer to move on the machine platform along a preset path.

[0027] First, after the wafer is fixed and stabilized on the machine, the motion mechanism drives the machine to move the wafer at a constant or variable speed along a preset path. In this embodiment, the method by which the motion mechanism drives the machine to move the wafer at a constant or variable speed along the preset path is not limited. For example, it can be to execute a preset work trajectory, or it can be achieved through a combination of conventional basic motion methods.

[0028] Based on the foregoing explanation and comparison Figure 2 It is known that although the wafer placement position is preset on the machine, due to the coordination between the robot arm and the motion machine, as well as the influence of other objective conditions, the actual placement position 202 of the wafer may not coincide with the predetermined position 201. This may include a placement position deviation 203, that is, the placement plane position deviates from the predetermined position 201, and / or, there may also be a wafer placement orientation angle deviation 204, that is, the wafer rotates during placement, causing the angle deviation.

[0029] S102. Start the scanning acquisition card to acquire the reflected light from the laser beam illuminating the wafer surface. Sample the wafer surface information from a preset starting point in the form of scattered points on the Archimedean spiral and cache it as a scanning image. The scanning image includes the edge image of the edge range of the scanned wafer.

[0030] Figure 5 This is a schematic diagram of the path of the Archimedean spiral. (Example:) Figure 5 As shown, the scanning acquisition card starts sampling from a preset starting point 501 on the outer edge contour of the wafer, and the sampling is performed according to the scattered point form on the Archimedean spiral, and the obtained scan image is cached. The sampling path is... Figure 5 The Archimedean spiral shown has a preset starting point 501 set at the beginning of the outermost loop of the spiral. From there, samples are taken along the spiral path in a scattered manner, ending at the center point of the innermost loop 503. Since the scanning acquisition card advances along the Archimedean spiral from the preset starting point 501 to capture images of the wafer, if the distance 502 between adjacent loops in the Archimedean spiral is too small, there will be too much overlap in the scanned images at the same location between adjacent loops, increasing the number of scanned images and the processing load for subsequent calculations. Conversely, if the distance 502 between adjacent loops in the Archimedean spiral is too large, there will be uncaptured areas on the wafer between the scanned images at the same location between adjacent loops, affecting the coverage of defect detection. Therefore, the distance 502 between adjacent loops needs to be determined by combining parameters such as the size of the scanned images captured by the scanning acquisition card, the scanning frequency, and the speed of the machine's uniform movement.

[0031] When the scanning acquisition card starts from the preset starting point 501 and sets a reasonable spacing between two adjacent turns 502 according to the Archimedes spiral, it captures images from the starting point 501 to the ending point 503. The acquired and cached scan images include the edge images of the edge range of the scanned wafer.

[0032] Reference Figure 3a and Figure 3b As shown, the edge image of the scanned wafer edge range includes a complete image covering the edge range information of the entire closed loop of the wafer, in order to Figure 3aFor example, the edge range can be understood as including all scanned images capturing the entire outer contour of the wafer, while the images scanned circle by circle inside the wafer up to endpoint 503, excluding the edge images, are the center images covering the central range of the wafer. Along as... Figure 5 The path of the Archimedes spiral continues to sample the center image of the wafer after the edge image has been sampled.

[0033] S103. When the wafer surface information of the edge image completely covers the edge range of the wafer, the calibration deviation value of the wafer is obtained according to the cached edge image, and the current position coordinates of the wafer are calibrated to the calibration position coordinates.

[0034] Because scatter sampling is based on, as follows Figure 5 The Archimedean spiral is shown, with a preset starting point 501 at the outermost starting point of the spiral and an ending point 503 at the center of the spiral, which is also the center of the wafer. Therefore, the sampled and cached scan image is initially an edge image including the edge range of the wafer. Of course, according to... Figure 5 As shown, after the edge image scanning is completed and cached, the center image of the wafer will continue to be sampled and cached.

[0035] When the wafer surface information in the edge image has completely covered the edge range of the wafer, meaning the wafer edge image has been sampled and cached, the wafer position calibration is performed simultaneously based on the cached edge image while continuing to sample and cache the wafer center image. Specifically, the cached edge image is processed and calculated to obtain the wafer calibration deviation value, and the current position coordinates of the wafer are calibrated to the calibration position coordinates based on the obtained calibration deviation value.

[0036] S104. Convert the scanned image and stitch the converted scanned images together to obtain the wafer scanned image under the calibration position coordinates.

[0037] After calibrating the wafer to the calibration position coordinates, the scanned image is transformed according to the calibration position coordinates. The transformed scanned images are then stitched together to obtain the scanned image of the wafer at the calibration position coordinates.

[0038] This application provides a wafer positioning and detection method and a positioning and detection system. The wafer positioning and detection method includes driving the wafer to move on a machine along a preset path; activating a scanning acquisition card to acquire reflected light from a laser beam irradiating the wafer surface; sampling wafer surface information from a preset starting point in a scattered manner along an Archimedean spiral and buffering it as a scanning image; the scanning image includes an edge image of the edge range of the scanned wafer; when the wafer surface information of the edge image completely covers the edge range of the wafer, obtaining the wafer calibration deviation value based on the buffered edge image, and calibrating the current position coordinates of the wafer to calibration position coordinates; converting the scanning image and stitching the converted scanning images together to obtain a wafer scanning image at the calibration position coordinates. When the wafer surface information of the edge image completely covers the edge range of the wafer, the cached edge image can be calculated and processed simultaneously during the scanning and caching process to obtain the wafer calibration deviation value. Based on the calibration deviation value, the current position coordinates of the wafer are calibrated to the calibration position coordinates. The cached scan image is then converted into a scan image under the calibration position coordinates, and the converted scan images are stitched together to obtain the wafer scan image under the calibration position coordinates. This enables wafer alignment correction and wafer defect detection during the sampling and caching process of the scan image. While completing the scan image, the wafer position calibration is completed and the output is a wafer scan image under the calibrated coordinate position, which significantly improves the efficiency of wafer alignment and wafer inspection.

[0039] In one feasible implementation of the embodiments of this application, such as Figure 6 As shown, S101, driving the wafer to move along a preset path on the machine includes: S1011. Fix the wafer on the machine.

[0040] S1012. Drive the machine to move linearly along the first direction and rotate around the geometric center of the wafer, wherein the first direction is any linear direction on the plane where the machine is located.

[0041] For example, first, the wafer is fixed on the machine tool. Then, the machine tool is driven to move linearly in a first direction, and the wafer fixed on the machine tool will move linearly with the machine tool. Simultaneously, the machine tool is driven to rotate around the geometric center of the wafer. In this way, the combination of linear motion and rotational motion realizes the uniform motion of the wafer along the Archimedean spiral. The distance 502 between two adjacent turns in the Archimedean spiral can also be determined by adjusting the speed of the linear motion and rotational motion respectively.

[0042] In one feasible implementation of this application's embodiments, step S102, which involves sampling wafer surface information from a preset starting point in the form of scattered points along an Archimedean spiral and caching it as a scanned image, includes: S1021. Starting from a preset starting point, sample points along an Archimedes spiral to obtain a sampled image containing wafer surface information at the current point location, and cache the sampled image.

[0043] S1022. Arrange the sampled images corresponding to the scattered points on each turn of the Archimedes spiral into a row, and then arrange multiple rows into columns to cache the sampled images as a two-dimensional Mercator diagram.

[0044] In step S102, during the process of sampling wafer surface information in the form of scattered points along the Archimedean spiral from a preset starting point, it is also necessary to arrange and stitch the sampled images of the scattered points. First, scattered images are sampled from the preset starting point along the Archimedean spiral towards the geometric center of the wafer to obtain a sampled image containing wafer surface information at the current scattered point position, and the sampled image is cached.

[0045] Then, taking each turn of the Archimedean spiral as a unit, each sample... Figure 1 The sampled images of the Archimedean spiral are arranged in a row, and the sampled images of adjacent spirals are arranged in columns. In this way, after sampling multiple spirals, the multiple rows and columns can be arranged to obtain a two-dimensional Mercator diagram.

[0046] In one feasible embodiment of this application, before S102, when starting the scanning acquisition card to acquire the reflected light irradiated by the laser beam on the wafer surface, the method further includes: S1020. Based on the current wafer size information and the pre-stored historical data information, determine the edge contour points of the current wafer, and use the edge contour points as the preset starting points.

[0047] Before activating the scanning and acquisition card, it is necessary to determine the preset starting point for sampling as accurately as possible. In this embodiment, this can be set based on past experience values. After each setting, the current value is also included in the experience value data to continuously correct and improve the accuracy of the experience value data.

[0048] Specifically, based on the current size information of the wafer to be inspected, the wafer radius is appropriately increased from the preset size information. For example, the existing wafer radius is increased by approximately 500 μm based on experience. Of course, to further set a more suitable increased size, the centrifugal distance can be calculated based on pre-stored historical alignment data.

[0049] Doffset= (1) Doffset is the centrifugal distance, DoffsetMean is the average centrifugal distance, and DoffsetSigma is the standard deviation of the centrifugal distance. The outward expansion distance Dexpand based on the wafer radius can then be calculated using the following formula: Dexpand = DoffsetMean + N DoffsetSigma(2) This allows the edge contour points of the current wafer to be determined, which can then be used as a preset starting point. After setting the preset starting point, discrete scan images are scanned on the wafer. When the wafer surface information of the edge image completely covers the edge range of the wafer, the edge image can fully characterize the edge feature information of the wafer.

[0050] In one feasible implementation of the embodiments of this application, the wafer surface information of the edge image completely covers the edge range of the wafer, including: the outer edge contour of the wafer includes a closed-loop adjacent round edge and a marked edge, the marked edge includes a flat cut edge or a V-shaped cut edge, the edge range of the wafer includes a round edge contour and a flat cut edge contour, or the edge range of the wafer includes a round edge contour and a V-shaped cut pattern.

[0051] like Figures 7a-7c As shown, the grid in the figure represents the area on the wafer covered by the scanned image. Figure 7a In the process, the scanned image covers an edge range of 701, resulting in a complete edge image. Figure 7b In the process, the scanned image covers the central area 702, thus obtaining the entire central image. Figure 7c In the process, the scanned image covers the entire wafer area 703, resulting in a complete wafer scan image. When the scanned area is as shown in the wafer area 703, it indicates that the scan of the entire wafer has been completed.

[0052] The outer edge contour of the wafer includes a closed-loop adjacent circular edge and a marking edge, such as... Figure 8a As shown, the marked edge includes the flat edge, and the edge range of the wafer includes the rounded edge contour and the flat edge contour. Therefore, the outer edge image needs to cover the rounded edge contour and the flat edge contour. Figure 8b As shown, the marked edge includes the V-shaped cut edge. Because the V-shaped cut edge has a V-shaped pattern structure, it extends deeper into the wafer than the flat cut edge. Therefore, the edge range of the wafer needs to include the rounded edge outline and the entire pattern range covering the V-shaped cut.

[0053] For example, when the wafer is Figure 8b When the structure shown includes a V-shaped notch edge, ensuring that the wafer surface information of the edge image completely covers the edge range of the wafer includes checking whether the distance of the current scan is greater than the depth of the V-shaped notch edge. This is done by scanning multiple times along the Archimedean spiral path until the desired depth is reached. Figure 7aWhen the grid coverage is in the middle range, the current distance of the Archimedes spiral scan is greater than the depth of the V-shaped notch edge. The scanned image has completely covered the edge range of the wafer. Continue scanning the center image of the wafer. At the same time, the cached edge image can be used for calculation processing and wafer calibration.

[0054] In one feasible implementation of this application embodiment, S103, when the wafer surface information of the edge image completely covers the edge range of the wafer, the wafer calibration deviation value is obtained according to the cached edge image, and the current position coordinates of the wafer are calibrated to the calibration position coordinates, including: S1031. Extract the circular edge pixels and the marked edge pixels in segments from the edge image.

[0055] S1032. Obtain the calibration deviation value of the wafer, wherein the center position of the wafer is obtained based on the edge pixels, and the orientation angle of the wafer is obtained based on the mark edge type and the mark edge pixels.

[0056] S1033. Adjust the coordinate position of the wafer according to the calibration deviation value, and / or adjust the coordinate system of the wafer according to the calibration deviation value to calibrate the current position coordinates of the wafer to the calibration position coordinates.

[0057] The process of calculating and processing the edge image to obtain the calibration deviation value and then calibrating the wafer coordinates based on this value includes: firstly, extracting circular edge pixels and marker edge pixels segmentally from the edge image; obtaining the wafer's center position information based on the circular edge pixels; and obtaining the wafer's orientation angle information based on the marker edge type and marker edge pixels. This process is used to calculate the wafer's calibration deviation value. Finally, adjusting the wafer's coordinate position and / or adjusting the wafer's coordinate system based on the calibration deviation value to calibrate the wafer's current position coordinates to the calibration position coordinates.

[0058] Specifically, the edge image is first preprocessed, including processing based on the type of the wafer being measured, such as whether the marked edge is a flat-cut edge or a V-shaped notch edge, and the number and length of flat-cut edges, etc. Figure 9 A partial schematic diagram of the wafer edge image shown. Figure 9 The image shown is a partial edge image of a wafer with a marked edge being a flat-cut edge, including arc edge sampling sub-image 801, flat-cut edge sampling sub-image 802, and arc edge sampling sub-image 803.

[0059] After integrating the edge image along the Y-axis and extracting the minimum value, the center of the cut edge is found. Based on the length of the cut edge, the following is extracted: Figure 9The edge image shown can be used to obtain the edge image representing the contour of the flat-cut edge from the sub-image. Further extraction of edge contour points yields the position information of the flat-cut edge. Since the sampling frequency is typically chosen to be 30MHz, 128 samples are taken from the outer edge of the wafer. With 1024 points, tens of thousands of location points can be obtained. Substituting these points into a straight line for fitting, the resulting data accuracy can reach the microradian level.

[0060] For the circular arc edge sampling sub-images 801 and 803, since the center data and edge image bandwidth have already been obtained in the edge image processing calculation of the flat-cut edge, the flat-cut edge information is retained proportionally for the circular arc edge to avoid introducing the boundary between the flat-cut edge and the circular arc edge. After extracting the edge contour, the position arrays of the two circular arc edges are obtained. After merging the arrays, there are approximately 800,000 points. By performing least squares fitting of a circle, the center and radius of the wafer can be obtained. Similarly, since the discrete point data of the circular arc edge can reach hundreds of thousands, the positioning accuracy of the wafer can reach the sub-micron level.

[0061] Calibrating wafer coordinate information based on calibration deviation values ​​can also include adjusting the wafer's coordinate position according to the calibration deviation value. This involves adjusting the wafer's position via machine movement to align its current coordinate position with a predetermined position, thereby eliminating positional deviations caused during wafer placement. Alternatively, it can involve adjusting the wafer's coordinate system based on the calibration deviation value to calibrate the wafer's current position coordinates to the calibration position coordinates. In other words, it's not necessary to actually move the wafer; the spatial coordinates of the wafer are calibrated within the processor to assign corrected coordinates to the scanned image.

[0062] For example, the calibration deviation values ​​(dx, dy, dθ) include the coordinate information of the marked edge and the coordinate information of the wafer.

[0063] Please refer to Figure 2 As shown, when there is a deviation between the actual placement position 202 and the predetermined position 201 of the wafer, the deviation typically includes a positional deviation 203 and an orientation angle deviation 204. The positional deviation 203 is denoted as (dx, dy), and the orientation angle deviation 204 is expressed as angle θ. Together, they constitute the calibration deviation value (dx, dy, dθ), which is used to characterize the deviation between the predetermined position 201 and the actual placement position 202.

[0064] In one feasible implementation of the embodiments of this application, S104, converting the scanned image and stitching the converted scanned image, includes: The Mercator image is reconstructed into a wafer image in Cartesian coordinates, where the wafer is mapped to an M-axis based on the wafer size and the pixel values ​​of the mosaic. The image contains M pixels. The pixel values ​​of the sub-image in row i and column j are calculated by bilinear interpolation of the corresponding scanned image in the Mercator image. The position of the sub-image in row i and column j is obtained by translating and rotating the scanned image and then transforming it into Cartesian coordinates. The sub-images are then stitched together to obtain the wafer image.

[0065] Once the wafer surface information in the edge image completely covers the edge range of the wafer, the scanning acquisition card continues to acquire the reflected light from the laser beam illuminating the wafer surface, continuing to acquire the center image of the wafer surface in the form of scattered points along an Archimedean spiral. Thus, the step of calibrating the wafer based on its calibration deviation value, and calibrating the wafer's current position coordinates to the calibration position coordinates, can be performed synchronously with the scanning and buffering process of the wafer's center image.

[0066] During the scanning image generation and caching process, a task thread is started synchronously to reconstruct the Mercator image from the sampled cache into a wafer image in Cartesian coordinates without positional errors using the deviation data.

[0067] The specific reconstruction process includes mapping the wafer to an M-shaped image based on the preset wafer outer diameter and mosaic pixel size. The pixel value at row i, column j of an M-pixel image is calculated using bilinear interpolation of the pixels in the Mercator image. The wafer position corresponding to point (i, j) is (X, Y), where X = i pixelSize, Y=j pixelSize.

[0068] There are two ways to calibrate the wafer from its current position coordinates (X, Y) to the calibration position coordinates: rotate first and then translate, or translate first and then rotate. The following explanation uses the rotation-then-translation method as an example.

[0069] XonWafer = X cos(θ) – Y sin(θ) + dx; (3) YonWafer = X sin(θ) –Y cos(θ) + dy; (4) R= (5) Convert to polar coordinates row = R / StepSize, column = Math.Atan2(Y, X); after obtaining the index (row, column) of (i, j) in the Mercator plot, use bilinear interpolation to calculate the pixel value of (i, j), and so on, to reconstruct the wafer image in Cartesian coordinates.

[0070] During the image acquisition and scanning process, the scanning range along the Y-axis is detected. When the scanning range is less than the remaining distance of the inner circle, the image acquisition of the entire wafer area can be considered complete, i.e., achieving the desired result. Figure 7c The wafer is shown in area 703. At this point, the image acquisition work of the scanning acquisition card can be ended, the machine movement can be stopped, and the wafer positioning and detection are completed.

[0071] Another aspect of this application provides a wafer positioning and detection system, including a wafer carrier stage, a motion mechanism mounted on the wafer carrier stage, a laser source, a receiving sensor, and a processor. The motion mechanism includes a horizontal motion platform that drives the wafer carrier stage to move in a planar manner and a rotation axis that drives the wafer carrier stage to rotate horizontally. The laser source emits a laser beam toward the wafer mounted on the wafer carrier stage, so the receiving sensor receives the laser beam reflected by the wafer to form a scan image of the current detection position on the wafer. The processor is electrically connected to the motion mechanism, the laser source, and the receiving sensor, and is used to control the operation of the laser source and the motion mechanism, and to buffer, convert, and stitch the scan image received by the receiving sensor.

[0072] Another aspect of this application provides a wafer positioning and inspection system, including a wafer carrier stage 402, a motion mechanism 403 that mounts the wafer carrier stage 402, a laser source 406, a receiving sensor 401, and a processor. The motion mechanism 403 includes a horizontal motion platform that drives the wafer carrier stage 402 to move in a planar manner and a rotation axis that drives the wafer carrier stage to rotate horizontally, thereby ensuring that the wafer moves under the drive of the wafer carrier stage 402, so that the scanning acquisition card samples the wafer surface information in the form of scattered points on the Archimedean spiral and caches it in the scanned image.

[0073] like Figure 10 As shown, the laser source 406 emits a laser beam toward the wafer under test 404 mounted on the wafer carrier stage 402. The laser beam is currently focused at the current sampling point 405. Therefore, the receiving sensor 401 receives the laser beam reflected from the current sampling point 405 of the wafer under test 404 to form a sampling map of the current sampling point 405 on the wafer under test 404. The processor ( Figure 10 (Not shown) is electrically connected to the motion mechanism 403, the laser source 406, and the receiving sensor 401, respectively, for controlling the operation of the laser source 406 and the motion mechanism 403, and for buffering, converting, and stitching the scanned image received by the receiving sensor 401. The receiving sensor 401 can be the aforementioned scanning acquisition card, used to sample wafer surface information and buffer it as a scanned image.

[0074] In one feasible implementation of this application, the processor extracts the edge image covering the edge range of the wafer 404 to be tested to obtain the calibration deviation value of the wafer, calculates the coordinate position of the wafer after calibration based on the calibration deviation value, and transforms and stitches the scanned image according to the coordinate position of the wafer after calibration to obtain the wafer scanned image under the calibration position coordinates.

[0075] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A wafer positioning and detection method, characterized in that, The method includes: Drive the wafer to move along a preset path on the machine; The scanning acquisition card is activated to acquire the reflected light from the laser beam illuminating the wafer surface. The wafer surface information is sampled from a preset starting point in the form of scattered points on the Archimedean spiral and cached as a scanning image. The scanning image includes the edge image of the edge range of the wafer. When the wafer surface information of the edge image completely covers the edge range of the wafer, the calibration deviation value of the wafer is obtained according to the cached edge image, and the current position coordinates of the wafer are calibrated to the calibration position coordinates. The scanned image is converted and then stitched together to obtain the wafer scanned image at the calibration position coordinates.

2. The wafer positioning and detection method according to claim 1, characterized in that, The movement of the driving wafer on the machine along a preset path includes: Fix the wafer on the machine; The machine is driven to move linearly along a first direction and rotate around the geometric center of the wafer, wherein the first direction is any linear direction on the plane where the machine is located.

3. The wafer positioning and detection method according to claim 1, characterized in that, The step of sampling wafer surface information from a preset starting point in the form of scattered points along an Archimedean spiral and caching it as a scanned image includes: Starting from a preset starting point, samples are taken in the form of scattered points along the Archimedes spiral to obtain a sampled image containing wafer surface information at the current scattered point position, and the sampled image is cached. The sampled images corresponding to the scattered points on each turn of the Archimedes spiral are arranged in a row, and multiple rows are arranged in columns to cache the sampled images as a two-dimensional Mercator diagram.

4. The wafer positioning and detection method according to any one of claims 1-3, characterized in that, Before the activation of the scanning acquisition card to acquire the reflected light from the laser beam irradiating the wafer surface, the method further includes: Based on the current wafer size information and pre-stored historical data information, the edge contour points of the current wafer are determined, and the edge contour points are used as the preset starting points.

5. The wafer positioning and detection method according to claim 1, characterized in that, The condition that the wafer surface information of the edge image completely covers the edge range of the wafer includes: The outer edge contour of the wafer includes a closed-loop adjacent round edge and a marked edge. The marked edge includes a flat-cut edge or a V-shaped notch edge. The edge range of the wafer includes a round edge contour and a flat-cut edge contour, or the edge range of the wafer includes a round edge contour and a V-shaped notch pattern.

6. The wafer positioning and detection method according to claim 5, characterized in that, The step of obtaining the calibration deviation value of the wafer based on the cached edge image and calibrating the current position coordinates of the wafer to the calibration position coordinates includes: Extract circular edge pixels and marked edge pixels in segments from the edge image; Obtain the calibration deviation value of the wafer, wherein the center position of the wafer is obtained based on the edge pixels, and the orientation angle of the wafer is obtained based on the mark edge type and the mark edge pixels; Adjust the coordinate position of the wafer according to the calibration deviation value, and / or adjust the coordinate system of the wafer according to the calibration deviation value, so as to calibrate the current position coordinates of the wafer to the calibration position coordinates.

7. The wafer positioning and detection method according to claim 3, characterized in that, The step of converting the scanned image and stitching the converted scanned images includes: The Mercator image is reconstructed into a wafer image in Cartesian coordinates, wherein the wafer is mapped to an M-axis based on the wafer size and the pixel values ​​of the mosaic. The image contains M pixels. The pixel values ​​of the sub-image in row i and column j are calculated by bilinear interpolation of the corresponding scanned image in the Mercator image. The position of the sub-image in row i and column j is obtained by translating and rotating the scanned image and then transforming it into Cartesian coordinates. The sub-images are then stitched together to obtain the wafer image.

8. The wafer positioning and detection method according to claim 1, characterized in that, The method further includes: When the wafer surface information of the edge image completely covers the edge range of the wafer, the scanning acquisition card continues to acquire the reflected light from the laser beam illuminating the wafer surface, and acquires the center image of the wafer surface in the form of scattered points on the Archimedean spiral.

9. A wafer positioning and inspection system, characterized in that, The device includes a wafer carrier stage, a motion mechanism mounted on the wafer carrier stage, a laser source, a receiving sensor, and a processor. The motion mechanism includes a horizontal motion platform that drives the wafer carrier stage to move in a planar manner and a rotation axis that drives the wafer carrier stage to rotate horizontally. The laser source emits a laser beam toward the wafer mounted on the wafer carrier stage, and the receiving sensor receives the laser beam reflected from the wafer to form a scan image of the current detection position on the wafer. The processor is electrically connected to the motion mechanism, the laser source, and the receiving sensor, and is used to control the operation of the laser source and the motion mechanism, and to buffer, convert, and stitch the scan image received by the receiving sensor.

10. The wafer positioning and inspection system according to claim 9, characterized in that, The processor extracts the edge image covering the edge range of the wafer to obtain the wafer's calibration deviation value, calculates the wafer's calibrated coordinate position based on the calibration deviation value, and transforms and stitches the scanned image based on the wafer's calibrated coordinate position to obtain the wafer scanned image at the calibrated position coordinates.

Images