Multi-region wafer reinspection method, device, equipment and storage medium

Abstract

The invention relates to the technical field of semiconductor detection, and discloses a multi-region wafer reinspection method, device and equipment and a storage medium, and the method comprises the steps: clustering initial inspection defects into a plurality of initial regions according to spatial distribution and according to initial inspection defect coordinates and process data of a wafer, and obtaining image features of the wafer; performing region boundary correction on the plurality of preliminary regions according to the image features to obtain a plurality of target regions; self-adaptively setting a region focus for each target region to adapt to the local height difference in the same region, and improving the reinspection precision according to each target region; and the shortest moving path of the camera during reinspection is calculated, the target areas are reinspected by using the corresponding area focuses according to the shortest moving path, the moving time of the mechanical platform and the Z-axis adjusting time are minimized, and the reinspection efficiency is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of semiconductor detection, and in particular to a multi-region wafer re-inspection method, device, equipment and storage medium. BACKGROUND

[0002] The depth of field of a high-power lens is very small (e.g. 20x lens depth of field ±1.5um), while the height difference of different functional areas (e.g. IDT, PAD, metal area) on the wafer surface is significant (up to 4um or more). The traditional global focusing or single-region focusing method needs to adjust the Z-axis repeatedly, resulting in low re-inspection efficiency (e.g. 100 defects require 100 times of focusing adjustment). The pre-stored focal point combined with the defect matching method in the prior art reduces the number of focusing times, but cannot adapt to local height fluctuations, resulting in low detection accuracy and long detection time.

[0003] Therefore, how to adapt to the local height difference in the same region, so as to balance the detection efficiency and detection accuracy, is a technical problem to be solved.

[0004] The above content is only used to assist in understanding the technical solutions of the present application, and does not represent the acknowledgement of the above content as prior art. SUMMARY

[0005] The main purpose of the present application is to provide a multi-region wafer re-inspection method, device, equipment and storage medium, which aims to solve the technical problem that the conventional method in the prior art cannot balance the detection efficiency while dynamically adapting to local height fluctuations, resulting in low detection efficiency and low detection quality.

[0006] To achieve the above purpose, the present application provides a multi-region wafer re-inspection method, which comprises the following steps: According to the wafer initial defect coordinates and process data, the initial defects are clustered into multiple preliminary regions according to the spatial distribution; Obtain the image features of the wafer, and correct the region boundaries of the multiple preliminary regions according to the image features to obtain multiple target regions; Adaptively set a region focal point for each target region; According to each target region, calculate the shortest moving path of the camera during re-inspection; According to the shortest moving path, use the corresponding region focal point to re-inspect each target region.

[0007] Preferably, after the re-inspection of each target region using the corresponding region focal point according to the shortest moving path, the method further comprises: Obtain the boundary regions between the target regions; Real-time fine-tuning focusing is performed on each boundary region; According to the focus after fine tuning, each of the boundary regions is re-inspected.

[0008] Preferably, the real-time fine tuning of each of the boundary regions comprises: According to the preset standard map, the position height value of each of the boundary regions is calculated; Taking the lowest height value in each of the position height values as a reference, the preset standard map is fitted as a curved surface by least squares; The inter-regional height difference of the boundary region is calculated in real time, and the boundary region image height is calculated according to the depth of field size; According to the boundary region image height, the focus after fine tuning is determined.

[0009] Preferably, the image features of the wafer are obtained, and the region boundary of the plurality of preliminary regions is corrected according to the image features to obtain a plurality of target regions, comprising: An image of the wafer is obtained; The edge enhancement or gray level mutation detection is performed on the image of the wafer to determine the true region boundary position of the plurality of preliminary regions; The true region boundary position is corrected by comparison and matching; The gray level mean and contrast of the corrected region are counted; According to the gray level mean and the contrast, the corrected region is classified to obtain a plurality of target regions.

[0010] Preferably, the target region is re-inspected according to the shortest moving path and the corresponding region focus, comprising: According to the region focus corresponding to each of the target regions, each of the target regions is subjected to one-time focusing by partitioning; The coordinates of the preliminary defects are obtained, and the target region where the preliminary defect is located is calculated according to the coordinates of the preliminary defect, and the corresponding region focus is selected as the re-inspection position of the preliminary defect; According to the shortest moving path and the re-inspection position, all of the preliminary defects of the wafer are re-inspected.

[0011] Preferably, before the region focus of each of the target regions is adaptively set, the method further comprises: A regression model of focus position and image features is established; The focus data of adjacent dies in the wafer is collected in real time, and the regression model is dynamically updated; Correspondingly, the region focus of each of the target regions is adaptively set, comprising: The region focus of each of the target regions is adaptively set by the regression model.

[0012] Preferably, the calculating the shortest moving path of the camera during re-inspection according to each of the target regions comprises: calculating the shortest moving path of the camera during re-inspection according to the platform acceleration, the Z-axis adjustment time, and the region priority of each of the target regions.

[0013] In addition, to achieve the above object, the present application further provides a multi-region wafer re-inspection device, which comprises: a clustering module, configured to cluster the preliminary regions according to the wafer initial inspection defect coordinates and process data; a correction module, configured to obtain image features of the wafer, and correct the region boundaries of the preliminary regions according to the image features to obtain target regions; a setting module, configured to set a region focus for each of the target regions; a calculating module, configured to calculate the shortest moving path of the camera during re-inspection according to each of the target regions; a re-inspection module, configured to re-inspect each of the target regions using the corresponding region focus according to the shortest moving path.

[0014] In addition, to achieve the above object, the present application further provides a multi-region wafer re-inspection device, which comprises a memory, a processor, and a multi-region wafer re-inspection program stored in the memory and executable on the processor, and the multi-region wafer re-inspection program is configured to implement the steps of the multi-region wafer re-inspection method as described above.

[0015] In addition, to achieve the above object, the present application further provides a storage medium, which stores a multi-region wafer re-inspection program, and the multi-region wafer re-inspection program is executable on a processor to implement the steps of the multi-region wafer re-inspection method as described above.

[0016] In the present application, the preliminary regions are clustered according to the wafer initial inspection defect coordinates and process data, the image features of the wafer are obtained, the region boundaries of the preliminary regions are corrected according to the image features to obtain target regions, a region focus is set for each of the target regions to adapt to the local height difference in the same region and improve the re-inspection accuracy, the shortest moving path of the camera during re-inspection is calculated according to each of the target regions, each of the target regions is re-inspected using the corresponding region focus according to the shortest moving path to minimize the mechanical platform moving and Z-axis adjustment time and improve the re-inspection efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0017] Figure 1This is a schematic diagram of the structure of a multi-region wafer re-inspection device in the hardware operating environment involved in the embodiments of the present invention; Figure 2 This is a flowchart illustrating the first embodiment of the multi-region wafer re-inspection method of the present invention; Figure 3 This is a structural block diagram of the first embodiment of the multi-region wafer re-inspection device of the present invention; Figure 4 This is a structural block diagram of the first embodiment of the multi-region wafer re-inspection system of the present invention.

[0018] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0019] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0020] Reference Figure 1 , Figure 1 This is a schematic diagram of the structure of a multi-region wafer re-inspection equipment in the hardware operating environment of an embodiment of the present invention.

[0021] like Figure 1 As shown, the multi-region wafer re-inspection equipment may include: a processor 1001, such as a central processing unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. The communication bus 1002 is used to enable communication between these components. The user interface 1003 may include a display screen, and optionally, it may also include a standard wired interface or a wireless interface. In this invention, the wired interface of the user interface 1003 may be a USB interface. The network interface 1004 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface). The memory 1005 may be a high-speed random access memory (RAM) or a non-volatile memory (NVM), such as a disk storage device. The memory 1005 may also optionally be a storage device independent of the aforementioned processor 1001.

[0022] Those skilled in the art will understand that Figure 1 The structure shown does not constitute a limitation on multi-region wafer re-inspection equipment and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0023] like Figure 1As shown, the memory 1005 as a computer storage medium can include an operating system, a network communication module, a user interface module, and a multi-region wafer review program.

[0024] In Figure 1 In the multi-region wafer review device shown, the network interface 1004 is mainly used to connect a background server and communicate data with the background server; the user interface 1003 is mainly used to connect a user device; the multi-region wafer review device calls the multi-region wafer review program stored in the memory 1005 through the processor 1001 and executes the multi-region wafer review method provided by the embodiment of the application.

[0025] Based on the above hardware structure, an embodiment of the multi-region wafer review method of the application is proposed.

[0026] Referring to Figure 2 , the first embodiment of the multi-region wafer review method of the application is proposed.

[0027] In the first embodiment, the multi-region wafer review method includes the following steps: Step S10: According to the wafer initial inspection defect coordinates and process data, the initial inspection defects are clustered into multiple preliminary regions according to the spatial distribution.

[0028] It should be noted that the execution subject of the embodiment is the multi-region wafer review device, wherein the multi-region wafer review device can be an electronic device such as a personal computer, a desktop computer, or a server, and the embodiment does not limit this. According to the wafer initial inspection defect coordinates and process data such as GDSII files, the position of each region on the Die is determined; the position of each region on the image can be directly framed and created, or the gray value difference of different regions can be extracted automatically through an algorithm, and DBSCAN or K-means algorithm is used to cluster the defects into IDT, PAD, and multiple preliminary regions of metal regions according to the spatial distribution.

[0029] Step S20: Obtain the image features of the wafer, correct the region boundaries of the multiple preliminary regions according to the image features, and obtain multiple target regions.

[0030] It can be understood that the image gray scale feature is combined to assist in the area classification, such as the high reflectivity of the metal area. Avoiding the area judgment of the defects one by one, the defects in the same target area can be directly batch processed. Based on the actual gray scale feature of the wafer image, the area defined in the process file (GDSII) is corrected. In particular, in the wafer detection, the GDSII file provides a theoretical layout, and the actual image often has a deviation, a distortion or a blurred boundary due to process deviation, etching deformation, alignment error and the like. According to the image feature of the wafer, the area edge of the plurality of preliminary areas is extracted. According to the extracted edge information, the area boundary is corrected through morphological boundary thinning combined with contrast matching, and a plurality of target areas are obtained.

[0031] Step S30: adaptively setting a region focus point for each target area.

[0032] It should be understood that the Z-axis is near the focus point, and the image definition is calculated within a certain range of movement. The point with the highest image definition is the focus point position, which is also the optimal Z-axis position. Specifically, taking the IDT area as an example: the camera moves above the IDT area for automatic focusing and photographing, and saves N field of view images. In order to prevent interference from other area features, the field of view images are cropped according to the center of the IDT area (assuming a rectangular area) and the width and height of the minimum bounding rectangle. The cropped N IDT area images are evaluated using a focusing algorithm to calculate the optimal Z-axis position and obtain the region focus point.

[0033] In a specific implementation, for each target area, a regression model between the region focus point position and the image features of each target area is established, the model is dynamically updated by real-time acquisition of the focus point data of adjacent dies, and the region focus point position is predicted and set. Thus, the height fluctuation in the same area is adapted, and the number of repeated focusing is reduced.

[0034] Step S40: calculating the shortest movement path of the camera during re-inspection according to each target area.

[0035] It should be understood that the platform acceleration, Z-axis adjustment time, and area priority (such as key defect priority) are calculated to obtain the shortest movement path of the camera to each target area during re-inspection, thereby reducing the time of defect re-inspection. Specifically, according to the genetic algorithm: some initial defects containing starting points and initial defects containing ending points are randomly generated as initial populations according to the platform acceleration, Z-axis adjustment time and area priority. The total distance of each path is calculated as fitness, the better path is taken as a mother node, and the paths in the mother node are randomly crossed to generate new individuals. The new individuals are randomly mutated to increase the diversity of the population. The above steps are repeated until the shortest movement path is found.

[0036] Step S50: According to the shortest moving path, each target area is re-inspected using the corresponding area focus.

[0037] In a specific implementation, for each target area after classification clustering, the target area is moved to the target area according to the planned shortest moving path, the mechanical platform moving time is minimized, and the predicted area focus is called to perform defect re-inspection. For the boundary area or the area focus with low confidence (for example, the model output confidence interval is large, or the prediction error historical value is greater than 1 μm), the area focus can be fine-tuned in real time, and the re-inspection accuracy can be improved.

[0038] In this embodiment, the wafer is classified into multiple preliminary areas according to the spatial distribution of the initial inspection defect coordinates and the process data of the wafer; image features of the wafer are obtained, the boundaries of the multiple preliminary areas are corrected according to the image features, and multiple target areas are obtained; a region focus is adaptively set for each target area; the shortest moving path of the camera during re-inspection is calculated according to each target area; and each target area is re-inspected using the corresponding area focus according to the shortest moving path. The mechanical platform moving and Z-axis adjustment time is minimized, the re-inspection efficiency is improved, the local height difference in the same area is adapted, and the re-inspection accuracy is improved.

[0039] Based on the first embodiment of the above method, further referring to Figure 2 , the second embodiment of the multi-area wafer re-inspection method of the present application is proposed.

[0040] In this embodiment, after the step S50, the following steps are further included: obtaining the boundary area between each target area; fine-tuning the focus of each boundary area in real time; re-inspecting each boundary area according to the focus after fine-tuning.

[0041] It should be noted that for the boundary area, for example, the area at the boundary of the IDT and the PAD, the real-time fine-tuning focus is used to balance the efficiency and the accuracy, and to avoid local blur.

[0042] Further, the real-time fine-tuning focus of each boundary area includes: calculating the position height value of each boundary area according to a preset standard map; using the minimum height value in the position height value as a reference, fitting the preset standard map to a curved surface by least squares; calculating the height difference between the boundary areas in real time, calculating the image height of the boundary area according to the depth of field; determining the focus after fine-tuning according to the image height of the boundary area.

[0043] It should be understood that, according to the preset standard map, the height value of each area position is calculated, the whole preset standard map is fitted as a curved surface based on the minimum height, the height of each point is calculated in real time according to the position, the height difference between the areas is calculated in real time in advance for the boundary area, and the height of the collected image of the boundary area is calculated according to the size of the depth of field.

[0044] Further, in the embodiment, the step S20 comprises: acquiring an image of the wafer; performing edge enhancement or gray level mutation detection on the image of the wafer to determine the real area boundary position of the plurality of preliminary areas; correcting the real area boundary position by comparison and matching to obtain a plurality of target areas.

[0045] It should be understood that the area defined in the process file (GDSII) is corrected based on the actual gray level features of the image. In wafer detection, the GDSII file provides a theoretical layout, and the actual image often has a shifted, distorted or blurred boundary due to process deviation, etching deformation, alignment error, etc.

[0046] It can be understood that the GDSII area is imported as a polygon or an area using image processing software; the actual image of the wafer is subjected to edge enhancement or gray level mutation detection to find the real area boundary position; the GDSII area and the image edge are compared, and the area is repositioned and matched by using morphological boundary thinning combined with comparison and matching, or a template matching method is used, in which a typical area in the actual image is used as a template for matching to accurately correct the original area position. The gray level mean value and contrast of the corrected area are counted for further classification and confirmation to determine whether it still meets the typical gray level characteristics of the IDT or PAD. In the embodiment, the correction of the real area boundary position by comparison and matching to obtain a plurality of target areas comprises: correcting the real area boundary position by comparison and matching; counting the gray level mean value and contrast of the corrected area; classifying the corrected area according to the gray level mean value and the contrast to obtain a plurality of target areas.

[0047] Further, in the embodiment, before the step S30, the method further comprises: establishing a regression model of the focus position and the image features; real-time acquisition of the focus data of the adjacent dies in the wafer to dynamically update the regression model; Correspondingly, the step S30 comprises: adaptively setting a region focus for each of the target areas by the regression model.

[0048] It should be noted that for each of the target regions, historical detection data such as the focal point positions of 10 adjacent dies are collected; and the regression model is trained: the input is the image features of the region such as texture, contrast and the like, and the output is the optimal Z-axis position, that is, the optimal region focal point is obtained.

[0049] Further, in the embodiment, the step S50 comprises: According to the platform acceleration, the Z-axis adjustment time, and the region priority of each of the target regions, the shortest moving path of the camera during re-inspection is calculated.

[0050] In a specific implementation, while improving the re-inspection efficiency, the detection accuracy is also taken into account, the platform acceleration, the Z-axis adjustment time, and the region priority of each of the target regions are taken as evaluation factors to calculate the shortest moving path of the camera during re-inspection. The region priority can be set according to the importance of defects, and important key defects are given priority.

[0051] In the embodiment, the boundary regions between the target regions are obtained, real-time fine focusing is performed on each of the boundary regions, and re-inspection is performed on each of the boundary regions according to the focal points after fine focusing. The boundary regions are fine-tuned in real time, so as to balance the efficiency and accuracy and avoid blur of the boundary regions.

[0052] In addition, the embodiment of the present application also provides a storage medium, and the storage medium stores a multi-region wafer re-inspection program. When the multi-region wafer re-inspection program is executed by a processor, the steps of the multi-region wafer re-inspection method described above are implemented.

[0053] In addition, with reference to Figure 3 , the embodiment of the present application also provides a multi-region wafer re-inspection device, which comprises: A clustering module 10 is configured to cluster the preliminary inspection defects into a plurality of preliminary regions according to the preliminary inspection defect coordinates and process data of the wafer. A correction module 20 is configured to obtain image features of the wafer, correct the region boundaries of the plurality of preliminary regions according to the image features, and obtain a plurality of target regions. A setting module 30 is configured to adaptively set a region focal point for each of the target regions. A calculation module 40 is configured to calculate a shortest moving path of a camera during re-inspection according to each of the target regions. A re-inspection module 50 is configured to use the corresponding region focal point to perform re-inspection on each of the target regions according to the shortest moving path.

[0054] Other embodiments or specific implementation manners of the multi-region wafer re-inspection device of the present application can refer to the above-mentioned method embodiments, and will not be described here again.

[0055] Further, based on the first embodiment of the above method, the embodiment of the present application further provides a multi-region wafer review system, referring to Figure 4 , the multi-region wafer review system comprises: a multi-region wafer review device 100, a camera 200, and a high-precision motion platform 300 as shown in Figure 1 The multi-region wafer review device 100 is connected with the camera 200 and the high-precision motion platform 300 respectively, the camera 200 is arranged on the high-precision motion platform 300, and the multi-region wafer review device 100 controls the camera 200 to collect images of a wafer to be detected. The wafer to be detected is usually placed on a chunk disc, the entire wafer is quickly scanned and the images of the wafer to be detected obtained by scanning are saved, the chunk disc is placed on the high-precision motion platform 300, and the multi-region wafer review device 100 controls the high-precision motion platform 300 to move to realize scanning of the entire wafer disc.

[0056] It should be noted that in this document, the terms "comprising", "containing", or any other variant thereof are intended to cover non-exclusive inclusions, so that a process, method, article, or system that includes a series of elements not only includes those elements, but also includes other elements not explicitly listed, or inherent to such a process, method, article, or system. Without more limitations, the element defined by the statement "comprising a" does not exclude the presence of another identical element in the process, method, article, or system that includes the element.

[0057] The above-mentioned embodiment numbers of the present application are only for description, and do not represent the advantages and disadvantages of the embodiments. In the unit claims of several devices, several of these devices can be embodied by the same hardware item. The use of the words first, second, and third does not indicate any order, and these words can be interpreted as identifiers.

[0058] From the above description of the embodiments, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be realized by means of software and a necessary general hardware platform, of course, they can also be realized by hardware, but in many cases, the former is a better embodiment. Based on such understanding, the technical solutions of the present application or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as a read-only memory image (ROM) / random access memory (RAM), a magnetic disk, an optical disk), and includes a plurality of instructions for making a terminal device (which can be a mobile phone, a computer, a server, an air conditioner, or a network device) execute the methods described in the embodiments of the present application.

[0059] The above merely provides the preferred embodiments of the present application, and is not intended to limit the patent scope of the present application. Any equivalent structure or equivalent flow transformation made according to the contents of the present application specification and drawings, or direct or indirect application in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A multi-zone wafer review method, comprising: The multi-region wafer review method comprises the following steps: According to the wafer inspection defect coordinates and process data, the wafer inspection defects are clustered into multiple preliminary regions according to the spatial distribution; Obtain the image features of the wafer, and correct the region boundaries of the multiple preliminary regions according to the image features to obtain multiple target regions; Adaptively set a region focus point for each target region; According to each target region, calculate the shortest moving path of the camera during review; According to the shortest moving path, review each target region using the corresponding region focus point.

2. The multi-zone wafer review method of claim 1, wherein, After the review of each target region using the corresponding region focus point according to the shortest moving path, the method further comprises the following steps: Obtain the boundary regions between the target regions; Real-time fine-tune the focus of each boundary region; Review each boundary region according to the focus after fine-tuning.

3. The multi-zone wafer review method of claim 2, wherein, The real-time fine-tune of the focus of each boundary region comprises the following steps: According to a preset standard image, calculate the position height value of each boundary region; Take the lowest height value in the position height values as a reference, and fit the preset standard image into a curved surface using the least squares method; Real-time calculate the height difference between regions of the boundary region, and calculate the image height of the boundary region according to the depth of field; Determine the focus after fine-tuning according to the image height of the boundary region.

4. The multi-zone wafer review method of claim 1, wherein, The method of obtaining the image features of the wafer, and correcting the region boundaries of the multiple preliminary regions according to the image features to obtain multiple target regions comprises the following steps: Obtain the image of the wafer; Determine the real region boundary position of the multiple preliminary regions by performing edge enhancement or gray level mutation detection on the image of the wafer; Correct the real region boundary position by comparison and matching; Statistically analyze the gray level mean value and contrast of the corrected region; Classify the corrected region according to the gray level mean value and the contrast to obtain multiple target regions.

5. The multi-zone wafer review method of claim 1, wherein, The review of each target region using the corresponding region focus point according to the shortest moving path comprises the following steps: According to the region focus point corresponding to each target region, perform one-time focusing on each target region; Obtain the coordinates of the wafer inspection defects, calculate the target region where the wafer inspection defects are located according to the coordinates of the wafer inspection defects, and select the corresponding region focus point as the review position of the wafer inspection defects; According to the shortest moving path and the review position, review all wafer inspection defects of the wafer.

6. The multi-zone wafer review method of claim 1, wherein, Before adaptively setting the region focus point for each target region, the method further comprises the following steps: Establish a regression model of the focus position and the image features; Real-time collect the focus data of adjacent dies in the wafer, and dynamically update the regression model; Correspondingly, the adaptively setting of the region focus point for each target region comprises the following steps: Adaptively set the region focus point for each target region through the regression model.

7. The multi-zone wafer review method of any one of claims 1-6, wherein, The calculation of the shortest moving path of the camera during review according to each target region comprises the following steps: According to the platform acceleration, the Z-axis adjustment time, and the region priority of each target region, calculate the shortest moving path of the camera during review.

8. A multi-zone wafer review apparatus, comprising: The multi-region wafer review device comprises: The clustering module is configured to cluster the initial inspection defects into a plurality of preliminary regions according to the initial inspection defect coordinates and the process data of the wafer; The correction module is configured to obtain image features of the wafer, and correct boundaries of the preliminary regions according to the image features to obtain a plurality of target regions; The setting module is configured to set a region focus point for each target region adaptively; The calculation module is configured to calculate a shortest moving path of a camera during re-inspection according to each target region; The re-inspection module is configured to re-inspect each target region according to the shortest moving path and using a corresponding region focus point.

9. A multi-zone wafer review apparatus, comprising: The multi-region wafer re-inspection device comprises a memory, a processor, and a multi-region wafer re-inspection program stored in the memory and executable on the processor, and the multi-region wafer re-inspection program is used to implement the steps of the multi-region wafer re-inspection method according to any one of claims 1 to 7 when executed by the processor.

10. A storage medium, characterized by The storage medium stores a multi-region wafer re-inspection program, and the multi-region wafer re-inspection program is used to implement the steps of the multi-region wafer re-inspection method according to any one of claims 1 to 7 when executed by the processor.

Images