Wafer defect detection method and device

Abstract

The invention provides a wafer defect detection method and device, and belongs to the technical field of semiconductor manufacturing. The wafer defect detection method comprises the steps that defect data of discontinuous defects on the surface of each wafer are acquired, and the defect data comprise the positions, sizes and types of the discontinuous defects; processing the defect data by using an image reconstruction algorithm, and reconstructing the discontinuous defect into a defect in a complete form; analyzing the spatial distribution characteristics of the defects in the complete form by utilizing a spatial clustering algorithm, and judging whether a preset defect distribution mode exists or not; if the preset defect distribution mode exists, calculating a risk score of the wafer according to the defect distribution density, the defect size and the defect distribution rule by utilizing a defect identification model; and comparing the risk score of the wafer with a preset risk threshold, and if the risk score is greater than the preset risk threshold, determining the wafer as an abnormal wafer. According to the invention, the discontinuous defect of the wafer can be detected.

Description

Technical Field

[0001] The present invention relates to the field of semiconductor manufacturing technology, and in particular to a wafer defect detection method and device. Background Art

[0002] With the continuous advancement of semiconductor manufacturing technology, defect control on the wafer surface has become particularly important. In the wafer manufacturing process, non-continuous defects (such as surface damage with a specific distribution or microscopic discontinuous defects) are an important reason affecting wafer quality. For defects with continuous length, they can generally be accurately judged by traditional detection systems and marked as NG (No Good) or G (Good). However, there is often great controversy in the judgment of non-continuous defects. The main reason is that the discontinuity and complex planar distribution characteristics of such defects make the standards based on manual judgment vague and easily affected by subjective factors, which not only reduces the accuracy of detection, but also may lead to misjudgment or missed detection. Summary of the invention

[0003] The present invention provides a wafer defect detection method and device, which can accurately detect non-continuous defects of the wafer.

[0004] In order to achieve the above object, the technical solution adopted by the embodiment of the present invention is:

[0005] A wafer defect detection method, comprising:

[0006] Acquire defect data of non-continuous defects on the surface of each wafer, wherein the defect data includes the location, size and type of the non-continuous defects;

[0007] Processing the defect data using an image reconstruction algorithm to reconstruct the discontinuous defect into a defect in a complete form;

[0008] Analyzing the spatial distribution characteristics of the defects in the complete form by using a spatial clustering algorithm to determine whether a preset defect distribution pattern exists;

[0009] If there is a preset defect distribution pattern, the risk score of the wafer is calculated using a defect recognition model according to the defect distribution density, the size of the defect and the distribution law of the defect;

[0010] The risk score of the wafer is compared with a preset risk threshold, and if the risk score is greater than the preset risk threshold, the wafer is determined to be an abnormal wafer.

[0011] In some embodiments, the preset defect distribution pattern includes at least one of the following:

[0012] Centralized distribution;

[0013] Edge clustering distribution;

[0014] Concentric circle distribution.

[0015] In some embodiments, before analyzing the spatial distribution characteristics of the defects in the complete form using a spatial clustering algorithm, the method further includes:

[0016] Judging whether the defect in the complete form meets a set defect standard according to the connectivity and length of the defect in the complete form;

[0017] If the defect meets the set defect standard, the step of analyzing the spatial distribution characteristics of the defect in the complete form by using a spatial clustering algorithm is performed.

[0018] In some embodiments, the method further comprises:

[0019] Matching the production process parameters of the wafer with the production process parameters of the first wafer in the historical database;

[0020] The risk threshold is adjusted according to a matching degree between a production process parameter of the wafer and a production process parameter of the first wafer, wherein the first wafer is determined to be an abnormal wafer.

[0021] In some embodiments, adjusting the risk threshold according to the matching degree between the production process parameters of the wafer and the production process parameters of the first wafer includes at least one of the following:

[0022] If the matching degree between the crystal growth rate curve of the wafer and the crystal growth rate curve of the first wafer is greater than or equal to a preset first matching degree, lowering the risk threshold; if the matching degree between the crystal growth rate curve of the wafer and the crystal growth rate curve of the first wafer is less than or equal to a preset second matching degree, raising the risk threshold;

[0023] If the matching degree between the crystal growth temperature curve of the wafer and the crystal growth temperature curve of the first wafer is greater than or equal to a preset first matching degree, the risk threshold is lowered; if the matching degree between the crystal growth temperature curve of the wafer and the crystal growth temperature curve of the first wafer is less than or equal to a preset second matching degree, the risk threshold is increased;

[0024] If the matching degree between the crystal growth pressure curve of the wafer and the crystal growth pressure curve of the first wafer is greater than or equal to a preset first matching degree, lowering the risk threshold; if the matching degree between the crystal growth pressure curve of the wafer and the crystal growth pressure curve of the first wafer is less than or equal to a preset second matching degree, raising the risk threshold;

[0025] The second matching degree is smaller than the first matching degree.

[0026] In some embodiments, after determining that the wafer is an abnormal wafer, the method further includes:

[0027] The wafer is intercepted, and a risk report of the wafer is generated, wherein the risk report includes defect modes, production process parameters and corresponding corrective measures of the wafer.

[0028] An embodiment of the present invention further provides a wafer defect detection device, comprising:

[0029] An acquisition module, used to acquire defect data of non-continuous defects on the surface of each wafer, wherein the defect data includes the location, size and type of the non-continuous defects;

[0030] A processing module, used to process the defect data using an image reconstruction algorithm to reconstruct the discontinuous defect into a defect of a complete form;

[0031] An analysis module, used to analyze the spatial distribution characteristics of the defects in the complete form by using a spatial clustering algorithm to determine whether a preset defect distribution pattern exists;

[0032] An identification module, for calculating the risk score of the wafer according to the defect distribution density, the size of the defect and the distribution law of the defect by using a defect identification model if a preset defect distribution pattern exists;

[0033] The determination module is used to compare the risk score of the wafer with a preset risk threshold, and if the risk score is greater than the preset risk threshold, determine the wafer as an abnormal wafer.

[0034] An embodiment of the present invention further provides an electronic device, comprising: a processor, a memory, and a program stored in the memory and executable on the processor, wherein the program, when executed by the processor, implements the steps of the wafer defect detection method as described above.

[0035] An embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the wafer defect detection method as described above are implemented.

[0036] An embodiment of the present invention further provides a computer program product, including computer instructions, which, when executed by a processor, implement the steps of the wafer defect detection method as described above.

[0037] The beneficial effects of the present invention are:

[0038] In this embodiment, the defect data of the non-continuous defects on the surface of each wafer is obtained, the defect data is processed using an image reconstruction algorithm, the non-continuous defects are reconstructed into defects of complete form, and the spatial distribution characteristics of the defects of complete form are analyzed using a spatial clustering algorithm. If there is a preset defect distribution pattern, the risk score of the wafer is calculated according to the defect distribution density, the size of the defect and the distribution law of the defect using a defect recognition model, and the risk score of the wafer is compared with the preset risk threshold to determine the abnormal wafer. This embodiment uses image reconstruction technology to connect and reconstruct the intermittent information of non-continuous defects into defects of complete form, which can effectively reduce the error of manual subjective judgment and ensure the accuracy of the detection results; this embodiment performs spatial distribution feature analysis based on spatial clustering, which can identify complex defect distribution patterns, especially high-risk defects with local aggregation or intermittent distribution, thereby reducing the possibility of missed detection and misjudgment; this embodiment can automatically identify abnormal wafers and intercept them, which can significantly improve the detection efficiency, effectively reduce manual intervention, and improve the recognition accuracy of complex defect patterns. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 A schematic diagram showing a flow chart of a wafer defect detection method according to an embodiment of the present invention;

[0040] Figure 2 A schematic diagram showing spatial clustering of defects according to an embodiment of the present invention;

[0041] Figure 3 A schematic diagram showing the structure of a wafer defect detection device according to an embodiment of the present invention;

[0042] Figure 4 A schematic diagram showing the structure of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0043] In order to make the purpose, technical solution and advantages of the embodiment of the present invention clearer, the technical solution of the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings of the embodiment of the present invention. Obviously, the described embodiment is a part of the embodiment of the present invention, not all of the embodiments. Based on the described embodiment of the present invention, all other embodiments obtained by ordinary technicians in this field belong to the scope of protection of the present invention.

[0044] The present invention provides a wafer defect detection method and device, which can detect batch defects of wafers.

[0045] The embodiment of the present invention provides a wafer defect detection method, such as Figure 1 As shown, including:

[0046] Step 101: Obtaining defect data of a discontinuous defect on the surface of each wafer, wherein the defect data includes the location, size and type of the discontinuous defect;

[0047] Specifically, the wafer surface particle detection equipment (SP5 / SP7) can be used to collect defect data of non-continuous defects on the surface of each wafer; and the collected defect data can be processed to eliminate noise in the defect data and errors in data collection, thereby improving the accuracy of defect detection.

[0048] Step 102: Processing the defect data using an image reconstruction algorithm to reconstruct the discontinuous defect into a defect of a complete form;

[0049] Specifically, discontinuous defect information can be reconstructed and connected into a complete form through edge detection and morphological processing techniques.

[0050] Step 103: Analyze the spatial distribution characteristics of the defects in the complete form by using a spatial clustering algorithm to determine whether there is a preset defect distribution pattern;

[0051] After defect image reconstruction, the density and distribution of defects on the wafer surface can be analyzed based on spatial distribution characteristics. If a specific form of defect aggregation appears in a certain area, such as local dense distribution, the area will be automatically marked as a risk area.

[0052] In this embodiment, a spatial clustering algorithm (such as DBSCAN or K-Means) can be used to analyze the discontinuous defects to determine whether the distribution pattern of the discontinuous defects in different regions conforms to the preset defect distribution pattern. Figure 2 A schematic diagram showing spatial clustering of defects according to an embodiment of the present invention, wherein the preset defect distribution pattern includes at least one of the following: central clustered distribution; edge clustered distribution; concentric circle band distribution.

[0053] In some embodiments, before analyzing the spatial distribution characteristics of the complete form defects using the spatial clustering algorithm, it can be determined whether the complete form defects meet the set defect standards based on the connectivity and length of the complete form defects; if they meet the set defect standards, the step of analyzing the spatial distribution characteristics of the complete form defects using the spatial clustering algorithm is performed. In this way, the connectivity and overall length of the non-continuous defects are first evaluated, and after meeting the set defect standards, the spatial distribution characteristics of the complete form defects are analyzed, which can reduce the amount of calculation of the spatial clustering algorithm.

[0054] Step 103: If there is a preset defect distribution pattern, a defect recognition model is used to calculate the risk score of the wafer according to the defect distribution density, the defect size and the defect distribution law;

[0055] In this embodiment, a large amount of historical detection data and data of abnormal wafers can be used in advance to train the defect recognition model using a machine learning algorithm. The input of the defect recognition model can be the defect distribution density, the size of the defect and the distribution law of the defect, and the output can be the risk score of the wafer. Among them, the risk score of the wafer is proportional to the size of the defect distribution density, and the risk score of the wafer is proportional to the size of the defect. If the distribution law of the defect conforms to a specific distribution law (such as central cluster distribution, edge cluster distribution, concentric circle band distribution), the risk score of the wafer is improved.

[0056] In order to improve the accuracy of detection, the number of defect recognition models can be multiple, and the multiple defect recognition models can be trained independently. The weight corresponding to each defect recognition model can be set according to the actual situation, and the detection results corresponding to the multiple defect recognition models are weighted and fused to obtain the final risk score. Among them, the multiple defect recognition models can be random forest models, decision tree models, deep learning neural network models, etc.

[0057] Step 105: Compare the risk score of the wafer with a preset risk threshold. If the risk score is greater than the preset risk threshold, determine the wafer as an abnormal wafer.

[0058] In this embodiment, the defect data of the non-continuous defects on the surface of each wafer is obtained, the defect data is processed using an image reconstruction algorithm, the non-continuous defects are reconstructed into defects of complete form, and the spatial distribution characteristics of the defects of complete form are analyzed using a spatial clustering algorithm. If there is a preset defect distribution pattern, the risk score of the wafer is calculated according to the defect distribution density, the size of the defect and the distribution law of the defect using a defect recognition model, and the risk score of the wafer is compared with the preset risk threshold to determine the abnormal wafer. This embodiment uses image reconstruction technology to connect and reconstruct the intermittent information of non-continuous defects into defects of complete form, which can effectively reduce the error of manual subjective judgment and ensure the accuracy of the detection results; this embodiment performs spatial distribution feature analysis based on spatial clustering, which can identify complex defect distribution patterns, especially high-risk defects with local aggregation or intermittent distribution, thereby reducing the possibility of missed detection and misjudgment; this embodiment can automatically identify abnormal wafers and intercept them, which can significantly improve the detection efficiency, effectively reduce manual intervention, and improve the recognition accuracy of complex defect patterns.

[0059] In order to improve the accuracy of the judgment, the production process parameters can be analyzed in association with historical data. In some embodiments, the method further includes:

[0060] Matching the production process parameters of the wafer with the production process parameters of the first wafer in the historical database;

[0061] The risk threshold is adjusted according to a matching degree between a production process parameter of the wafer and a production process parameter of the first wafer, wherein the first wafer is determined to be an abnormal wafer.

[0062] In some embodiments, adjusting the risk threshold according to the matching degree between the production process parameters of the wafer and the production process parameters of the first wafer includes at least one of the following:

[0063] If the matching degree between the crystal growth rate curve of the wafer and the crystal growth rate curve of the first wafer is greater than or equal to a preset first matching degree, lowering the risk threshold; if the matching degree between the crystal growth rate curve of the wafer and the crystal growth rate curve of the first wafer is less than or equal to a preset second matching degree, raising the risk threshold;

[0064] If the matching degree between the crystal growth temperature curve of the wafer and the crystal growth temperature curve of the first wafer is greater than or equal to a preset first matching degree, the risk threshold is lowered; if the matching degree between the crystal growth temperature curve of the wafer and the crystal growth temperature curve of the first wafer is less than or equal to a preset second matching degree, the risk threshold is increased;

[0065] If the matching degree between the crystal growth pressure curve of the wafer and the crystal growth pressure curve of the first wafer is greater than or equal to a preset first matching degree, lowering the risk threshold; if the matching degree between the crystal growth pressure curve of the wafer and the crystal growth pressure curve of the first wafer is less than or equal to a preset second matching degree, raising the risk threshold;

[0066] The second matching degree is smaller than the first matching degree.

[0067] The production process parameters of the wafer include the crystal growth rate curve, crystal growth temperature curve and crystal growth pressure curve corresponding to the wafer, wherein the crystal growth rate curve can characterize the speed change during the crystal growth process, the crystal growth temperature curve can characterize the ambient temperature change during the crystal growth process, and the crystal growth pressure curve can characterize the ambient pressure change during the crystal growth process. By judging the matching degree between the crystal growth pressure curve of the wafer and the crystal growth pressure curve of the first wafer, the matching degree between the crystal growth temperature curve of the wafer and the crystal growth temperature curve of the first wafer, and the matching degree between the crystal growth rate curve of the wafer and the crystal growth rate curve of the first wafer, the matching degree between the production process parameters of the wafer and the production process parameters of the first wafer can be judged. If the matching degree between the production process parameters of the wafer and the production process parameters of the first wafer is high, the risk threshold can be lowered; if the matching degree between the production process parameters of the wafer and the production process parameters of the first wafer is high, the risk threshold can be increased. In this embodiment, by combining historical data, the risk threshold can be dynamically updated according to different process conditions to ensure that a high interception accuracy rate can be maintained under different process conditions, reduce excessive detection of minor or no-effect defects, and reduce the false alarm rate of wafers, thereby improving production efficiency and yield rate.

[0068] In this embodiment, potential process problems or equipment abnormalities can be identified based on the determination results of the wafer. If a wafer is determined to be an abnormal wafer, it indicates that the production process parameters corresponding to the wafer (such as crystal growth rate, temperature, pressure, etc.) have potential process problems or equipment abnormalities. In some embodiments, after determining that the wafer is an abnormal wafer, the method further includes:

[0069] The wafer is intercepted, and a risk report of the wafer is generated, wherein the risk report includes defect modes, production process parameters and corresponding corrective measures of the wafer.

[0070] This embodiment can intercept abnormal wafers in time and provide detailed risk reports and improvement suggestions for manual review, which can ensure the accuracy of detection and reduce errors.

[0071] Based on the results of image reconstruction and morphological analysis, this embodiment can automatically determine whether defects need to be intercepted and generate risk reports for further verification, thereby improving the quality control level of wafers. This embodiment not only improves the accuracy of non-continuous defect detection, but also effectively intercepts defects with special clustered distribution, thereby improving the inspection efficiency and quality control level before wafer shipment.

[0072] The embodiment of the present invention also provides a wafer defect detection device, such as Figure 3 As shown, including:

[0073] An acquisition module 21 is used to acquire defect data of a non-continuous defect on the surface of each wafer, wherein the defect data includes the location, size and type of the non-continuous defect;

[0074] A processing module 22 is used to process the defect data using an image reconstruction algorithm to reconstruct the discontinuous defect into a defect of a complete form;

[0075] An analysis module 23 is used to analyze the spatial distribution characteristics of the defects in the complete form by using a spatial clustering algorithm to determine whether there is a preset defect distribution pattern;

[0076] An identification module 24 is used to calculate the risk score of the wafer according to the defect distribution density, the size of the defect and the distribution law of the defect by using a defect identification model if a preset defect distribution pattern exists;

[0077] The determination module 25 is used to compare the risk score of the wafer with a preset risk threshold, and if the risk score is greater than the preset risk threshold, determine the wafer as an abnormal wafer.

[0078] In this embodiment, the defect data of the non-continuous defects on the surface of each wafer is obtained, the defect data is processed using an image reconstruction algorithm, the non-continuous defects are reconstructed into defects of complete form, and the spatial distribution characteristics of the defects of complete form are analyzed using a spatial clustering algorithm. If there is a preset defect distribution pattern, the risk score of the wafer is calculated according to the defect distribution density, the size of the defect and the distribution law of the defect using a defect recognition model, and the risk score of the wafer is compared with the preset risk threshold to determine the abnormal wafer. This embodiment uses image reconstruction technology to connect and reconstruct the intermittent information of non-continuous defects into defects of complete form, which can effectively reduce the error of manual subjective judgment and ensure the accuracy of the detection results; this embodiment performs spatial distribution feature analysis based on spatial clustering, which can identify complex defect distribution patterns, especially high-risk defects with local aggregation or intermittent distribution, thereby reducing the possibility of missed detection and misjudgment; this embodiment can automatically identify abnormal wafers and intercept them, which can significantly improve the detection efficiency, effectively reduce manual intervention, and improve the recognition accuracy of complex defect patterns.

[0079] In some embodiments, the preset defect distribution pattern includes at least one of the following:

[0080] Centralized distribution;

[0081] Edge clustering distribution;

[0082] Concentric circle distribution.

[0083] In some embodiments, the processing module 22 is further used to determine whether the defect in the complete form meets the set defect standard according to the connectivity and length of the defect in the complete form;

[0084] If the defect meets the set defect standard, the analysis module 23 is used to perform a step of analyzing the spatial distribution characteristics of the defect in the complete form by using a spatial clustering algorithm.

[0085] In some embodiments, the determination module 25 is also used to match the production process parameters of the wafer with the production process parameters of the first wafer in the historical database; adjust the risk threshold according to the degree of matching between the production process parameters of the wafer and the production process parameters of the first wafer, wherein the first wafer is determined to be an abnormal wafer.

[0086] In some embodiments, the determination module 25 is specifically configured to perform at least one of the following:

[0087] If the matching degree between the crystal growth rate curve of the wafer and the crystal growth rate curve of the first wafer is greater than or equal to a preset first matching degree, lowering the risk threshold; if the matching degree between the crystal growth rate curve of the wafer and the crystal growth rate curve of the first wafer is less than or equal to a preset second matching degree, raising the risk threshold;

[0088] If the matching degree between the crystal growth temperature curve of the wafer and the crystal growth temperature curve of the first wafer is greater than or equal to a preset first matching degree, the risk threshold is lowered; if the matching degree between the crystal growth temperature curve of the wafer and the crystal growth temperature curve of the first wafer is less than or equal to a preset second matching degree, the risk threshold is increased;

[0089] If the matching degree between the crystal growth pressure curve of the wafer and the crystal growth pressure curve of the first wafer is greater than or equal to a preset first matching degree, lowering the risk threshold; if the matching degree between the crystal growth pressure curve of the wafer and the crystal growth pressure curve of the first wafer is less than or equal to a preset second matching degree, raising the risk threshold;

[0090] The second matching degree is smaller than the first matching degree.

[0091] In some embodiments, the device further comprises:

[0092] The interception module is used to intercept the abnormal wafer and generate a risk report of the wafer, wherein the risk report includes the defect mode, production process parameters and corresponding corrective measures of the wafer.

[0093] Please refer to Figure 4An embodiment of the present invention further provides an electronic device 30, including a processor 31, a memory 32, and a computer program stored in the memory 32 and executable on the processor 31. When the computer program is executed by the processor 31, each process of the above-mentioned wafer defect detection method embodiment is implemented, and the same technical effect can be achieved. To avoid repetition, it will not be described here.

[0094] The embodiment of the present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, each process of the above-mentioned wafer defect detection method embodiment is implemented, and the same technical effect can be achieved. To avoid repetition, it is not repeated here. Among them, the computer-readable storage medium includes permanent and non-permanent, removable and non-removable media, and information storage can be realized by any method or technology. The information can be a computer-readable instruction, a data structure, a module of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, read-only compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette, magnetic tape magnetic disk storage or other magnetic storage terminal device to be detected or any other non-transmission medium, which can be used to store information that can be accessed by the terminal device to be detected. According to the definition in this article, computer-readable storage media does not include temporary computer-readable media (transitory media), such as modulated data signals and carrier waves.

[0095] The present application also provides a computer program product, including computer instructions, which, when executed by a processor, implement the above Figure 1 The various processes of the method embodiment shown can achieve the same technical effect, and will not be described again here to avoid repetition.

[0096] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element.

[0097] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, 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 ROM / RAM, a magnetic disk, or an optical disk), and includes a number of instructions for enabling a terminal (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to execute the methods described in each embodiment of the present invention.

[0098] In the various method embodiments of the present disclosure, the serial numbers of the steps cannot be used to limit the sequence of the steps. For ordinary technicians in this field, without paying any creative work, changes to the sequence of the steps are also within the protection scope of the present disclosure.

[0099] It should be noted that each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the embodiments, since they are basically similar to the product embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the product embodiments.

[0100] The above is a preferred embodiment of the present disclosure. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles described in the present disclosure. These improvements and modifications should also be regarded as the scope of protection of the present disclosure.

Claims

1. A wafer defect detection method, characterized in that: include: Acquire defect data of non-continuous defects on the surface of each wafer, wherein the defect data includes the location, size and type of the non-continuous defects; Processing the defect data using an image reconstruction algorithm to reconstruct the discontinuous defect into a defect in a complete form; Analyzing the spatial distribution characteristics of the defects in the complete form by using a spatial clustering algorithm to determine whether a preset defect distribution pattern exists; If there is a preset defect distribution pattern, the risk score of the wafer is calculated using a defect recognition model according to the defect distribution density, the size of the defect and the distribution law of the defect; The risk score of the wafer is compared with a preset risk threshold, and if the risk score is greater than the preset risk threshold, the wafer is determined to be an abnormal wafer.

2. The wafer defect detection method according to claim 1, characterized in that: The preset defect distribution mode includes at least one of the following: Centralized distribution; Edge clustering distribution; Concentric circle distribution.

3. The wafer defect detection method according to claim 1, characterized in that: Before analyzing the spatial distribution characteristics of the defects in the complete form by using a spatial clustering algorithm, the method further includes: Judging whether the defect in the complete form meets a set defect standard according to the connectivity and length of the defect in the complete form; If the defect meets the set defect standard, the step of analyzing the spatial distribution characteristics of the defect in the complete form by using a spatial clustering algorithm is performed.

4. The wafer defect detection method according to claim 3, characterized in that: The method further comprises: Matching the production process parameters of the wafer with the production process parameters of the first wafer in the historical database; The risk threshold is adjusted according to a matching degree between a production process parameter of the wafer and a production process parameter of the first wafer, wherein the first wafer is determined to be an abnormal wafer.

5. The wafer defect detection method according to claim 4, characterized in that: The adjusting the risk threshold according to the matching degree between the production process parameters of the wafer and the production process parameters of the first wafer comprises at least one of the following: If the matching degree between the crystal growth rate curve of the wafer and the crystal growth rate curve of the first wafer is greater than or equal to a preset first matching degree, lowering the risk threshold; if the matching degree between the crystal growth rate curve of the wafer and the crystal growth rate curve of the first wafer is less than or equal to a preset second matching degree, raising the risk threshold; If the matching degree between the crystal growth temperature curve of the wafer and the crystal growth temperature curve of the first wafer is greater than or equal to a preset first matching degree, the risk threshold is lowered; if the matching degree between the crystal growth temperature curve of the wafer and the crystal growth temperature curve of the first wafer is less than or equal to a preset second matching degree, the risk threshold is increased; If the matching degree between the crystal growth pressure curve of the wafer and the crystal growth pressure curve of the first wafer is greater than or equal to a preset first matching degree, lowering the risk threshold; if the matching degree between the crystal growth pressure curve of the wafer and the crystal growth pressure curve of the first wafer is less than or equal to a preset second matching degree, raising the risk threshold; The second matching degree is smaller than the first matching degree.

6. The wafer defect detection method according to claim 1, characterized in that: After determining that the wafer is an abnormal wafer, the method further includes: The wafer is intercepted, and a risk report of the wafer is generated, wherein the risk report includes defect modes, production process parameters and corresponding corrective measures of the wafer.

7. A wafer defect detection device, characterized in that: include: An acquisition module, used to acquire defect data of non-continuous defects on the surface of each wafer, wherein the defect data includes the location, size and type of the non-continuous defects; A processing module, used to process the defect data using an image reconstruction algorithm to reconstruct the discontinuous defect into a defect of a complete form; An analysis module, used to analyze the spatial distribution characteristics of the defects in the complete form by using a spatial clustering algorithm to determine whether a preset defect distribution pattern exists; An identification module, for calculating the risk score of the wafer according to the defect distribution density, the size of the defect and the distribution law of the defect by using a defect identification model if a preset defect distribution pattern exists; The determination module is used to compare the risk score of the wafer with a preset risk threshold, and if the risk score is greater than the preset risk threshold, determine the wafer as an abnormal wafer.

8. An electronic device, characterized in that: include: A processor, a memory, and a program stored in the memory and executable on the processor, wherein when the program is executed by the processor, the steps of the wafer defect detection method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the wafer defect detection method according to any one of claims 1 to 6 are implemented.

10. A computer program product, characterized in that The method comprises computer instructions, which, when executed by a processor, implement the steps of the wafer defect detection method according to any one of claims 1 to 6.

Images