Wafer defect detection method, device and equipment

Abstract

The invention provides a wafer defect detection method, device and equipment, and relates to the technical field of semiconductors, the wafer defect detection method comprises the following steps: scanning a first wafer to obtain a defect distribution diagram of the first wafer; analyzing the defect distribution diagram to obtain defect feature information of the first wafer, the defect feature information comprising at least one of the following items: first defect density information and first cluster defect information; and intercepting the first wafer according to the defect feature information and the interception index. According to the embodiment of the invention, the potential abnormity which cannot be reflected by the overall defect density of the wafer can be detected through the first defect density information and / or the first cluster defect information, and the wafer can be effectively intercepted based on the first defect density information and / or the first cluster defect information and the corresponding interception index.

Description

Technical Field

[0001] The present invention relates to the field of semiconductor technology, and particularly to a method, device and equipment for detecting wafer defects. Background Art

[0002] During the manufacturing process of wafers, for intercepting products with cluster-like aggregation defects, the currently used interception methods usually use a wafer surface particle counter (such as SP5 equipment / SP7 equipment) to scan the wafer surface to obtain the position information of the defects, generate a defect distribution map based on the position information of the defects generated by the scan, and perform density analysis on the defect distribution map with an artificial or automatic defect classification (ADC) system. According to the number of defects and the size of the wafer, calculate the overall defect density of the wafer, compare the defect density with the limit sample, and intercept the wafer according to the comparison result.

[0003] However, for some abnormal cluster-like aggregations, especially when it is difficult to quantify the size and distribution type of the aggregation area, the existing methods are difficult to accurately intercept. However, these special aggregation area distributions may reflect potential defect risks, and under the existing rules, it is impossible to effectively intercept these wafers with risks. Summary of the Invention

[0004] Embodiments of the present invention provide a method, device and equipment for detecting wafer defects to solve the problem in the prior art that the wafer defects cannot be accurately detected, resulting in the inability to effectively intercept the risky wafers.

[0005] To solve the above technical problems, the embodiments of the present invention provide the following technical solutions:

[0006] In a first aspect, an embodiment of the present invention provides a method for detecting wafer defects, including:

[0007] Scanning a first wafer to obtain a defect distribution map of the first wafer;

[0008] Analyzing the defect distribution map to obtain defect feature information of the first wafer, where the defect feature information includes at least one of the following: first defect density information, first cluster defect information;

[0009] Intercepting the first wafer according to the defect feature information and the interception index.

[0010] In some embodiments, analyzing the defect distribution map to obtain the defect feature information of the first wafer includes:

[0011] Dividing the surface of the first wafer into multiple local regions;

[0012] Analyze the defect distribution map using a spatial clustering algorithm to obtain the local defect density of each local area.

[0013] Obtain the first defect density information of the first wafer according to the local defect density of each local area.

[0014] In some embodiments, the surface of the first wafer is divided into multiple local areas, including:

[0015] Taking the center of the surface of the first wafer as the center, a circular local area is divided in the central area of the surface of the first wafer.

[0016] Divide the peripheral area of the circular local area into multiple annular local areas.

[0017] Among them, the multiple annular local areas are concentrically arranged, and the radius of the circular local area is equal to the annular radius of each annular local area.

[0018] In some embodiments, intercepting the first wafer according to the defect feature information and the interception index includes:

[0019] Determine to intercept the first wafer when the first condition is met.

[0020] Among them, the first condition includes at least one of the following:

[0021] The first defect density information indicates that the local defect density of the local area of the first wafer exceeds the threshold density range.

[0022] The first defect density information indicates that the density difference between the local defect density of the first local area and the local defect density of the second local area of the first wafer is greater than the preset density difference, where the first local area is one of the multiple local areas on the surface of the first wafer, and the second local area is one of the multiple local areas on the surface of the first wafer other than the first local area.

[0023] In some embodiments, the method further includes:

[0024] Obtain the second defect density information of the second wafer.

[0025] Perform regression analysis on the second defect density information to obtain the threshold density range.

[0026] Among them, the second wafer is a wafer with qualified quality in the previous batch of the first wafer.

[0027] In some embodiments, the defect distribution map is analyzed to obtain the defect characteristic information of the first wafer, including:

[0028] Feature extraction is performed on the defect distribution map to obtain the first cluster defect information of the first wafer;

[0029] Wherein, the first cluster defect information includes at least one of the following:

[0030] The morphological characteristics of the cluster defects on the first wafer;

[0031] The scale quantity of the cluster defects on the first wafer;

[0032] The distribution size of the cluster defects on the first wafer;

[0033] The distribution density of the cluster defects on the first wafer.

[0034] In some embodiments, the first wafer is intercepted according to the defect characteristic information and the interception index, including:

[0035] The first wafer is intercepted when a second condition is satisfied;

[0036] Wherein, the second condition includes at least one of the following:

[0037] The morphological characteristics of the cluster defects on the first wafer indicated by the first cluster defect information conform to the morphological characteristics of the cluster defects on the third wafer, wherein the third wafer is a wafer with unqualified quality in a previous batch of the first wafer;

[0038] The scale quantity of the cluster defects on the first wafer indicated by the first cluster defect information is greater than a threshold quantity, and the threshold quantity is determined according to the scale quantity of the cluster defects on the fourth wafer;

[0039] The distribution size of the cluster defects on the first wafer indicated by the first cluster defect information is greater than a threshold size, and the threshold size is determined according to the distribution size of the cluster defects on the fourth wafer;

[0040] The distribution density of the cluster defects on the first wafer indicated by the first cluster defect information is greater than a threshold density, and the threshold density is determined according to the distribution density of the cluster defects on the fourth wafer;

[0041] Wherein, the fourth wafer is a wafer with qualified quality in a previous batch of the first wafer.

[0042] In some embodiments, the morphological characteristics of the cluster defects include at least one of the following:

[0043] The cluster defects exhibit linear characteristics;

[0044] The cluster defect has an annular feature;

[0045] The cluster defect has a local cluster feature.

[0046] In a second aspect, an embodiment of the present invention further provides a wafer defect detection device, including:

[0047] A first processing module, configured to scan a first wafer to obtain a defect distribution map of the wafer;

[0048] A second processing module, configured to analyze the defect distribution map to obtain defect feature information of the first wafer, where the defect feature information includes at least one of the following: first defect density information, first cluster defect information;

[0049] A third processing module, configured to intercept the first wafer according to the defect feature information and an interception index.

[0050] In a third aspect, an embodiment of the present invention further provides a wafer defect detection device, including: a processor, a memory, and a program stored on the memory and executable on the processor, where when the program is executed by the processor, the steps in the wafer defect detection method according to any one of the first aspects are implemented.

[0051] The embodiments of the present invention have the following beneficial effects:

[0052] The wafer defect detection method provided by the embodiment of the present invention scans a first wafer to obtain a defect distribution map of the first wafer, analyzes the defect distribution map to obtain defect feature information of the first wafer, where the defect feature information includes first defect density information and / or first cluster defect information. Potential anomalies that cannot be reflected by the overall defect density of the wafer can be detected through the first defect density information and / or the first cluster defect information. Based on the first defect density information and / or the first cluster defect information and the corresponding interception index, the first wafer is intercepted. Compared with the method of intercepting based on the overall defect density, the embodiment of the present invention can accurately detect the defect features on the wafer, and based on the defect features, the wafer can be effectively intercepted. Description of the Drawings

[0053] Figure 1 A flowchart showing the wafer defect detection method provided by the embodiment of the present invention;

[0054] Figure 2 A schematic diagram showing a partial area divided on the surface of the first wafer provided by the embodiment of the present invention;

[0055] Figure 3One of the schematic diagrams showing the defect distribution on the first wafer provided by the embodiments of the present invention;

[0056] Figure 4 Another schematic diagram showing the defect distribution on the first wafer provided by the embodiments of the present invention;

[0057] Figure 5 Schematic diagram of the structure of the wafer defect detection device provided by the embodiments of the present invention;

[0058] Figure 6 Schematic diagram of the structure of the wafer defect detection equipment provided by the embodiments of the present invention. Detailed implementation manners

[0059] To make the technical problems, technical solutions and advantages to be solved by the embodiments of the present invention clearer, the following will be described in detail with reference to the drawings and specific embodiments.

[0060] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the drawings of the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the described embodiments of the present invention fall within the scope of protection of the present invention.

[0061] To solve the problem in the prior art that the wafer defects cannot be accurately detected, resulting in the inability to effectively intercept risky wafers, the embodiments of the present invention provide a wafer defect detection method, device and equipment.

[0062] As Figure 1 shown, the embodiments of the present invention provide a wafer defect detection method, including:

[0063] Step 101: Scan the first wafer to obtain a defect distribution map of the first wafer.

[0064] In this step, a particle counter is used to scan the first wafer to obtain a defect distribution map (Defect distribution map) of the first wafer, and this defect distribution map is used to indicate the defect distribution at different positions on the first wafer.

[0065] Step 102: Analyze the defect distribution map to obtain defect feature information of the first wafer, where the defect feature information includes at least one of the following: first defect density information, first cluster defect information.

[0066] In this step, by performing defect density analysis on the defect distribution map, the first defect density information of the first wafer is obtained, and / or by performing defect distribution characteristic analysis on the defect distribution map, the first cluster defect information of the first wafer is obtained.

[0067] Among them, the first defect density information is used to indicate the defect density conditions at different positions on the first wafer. The first cluster defect information is used to indicate the morphological information and distribution position information of the cluster defects at different positions on the first wafer.

[0068] The first defect density information and / or the first cluster defect information can more accurately reflect the distribution of defects on the first wafer and the potential abnormalities of the defects. For example, there are obvious differences in the defect aggregation degree between some position regions and other position regions, or the aggregation degree of cluster-shaped defects exceeds a certain number, or the size of the cluster-shaped defects is larger than a certain regional size.

[0069] Step 103: Intercept the first wafer according to the defect characteristic information and the interception index.

[0070] After obtaining the first defect density information of the first wafer and / or the first cluster defect information of the first wafer, corresponding interception indexes are set. When the first defect density information of the first wafer meets the corresponding interception index and / or the first cluster defect information of the first wafer meets the corresponding interception index, it is determined that the first wafer does not meet the shipping requirements or the quality of the first wafer is unqualified and needs to be intercepted, which improves the efficiency and accuracy of shipping interception.

[0071] In some embodiments, analyzing the defect distribution map to obtain the defect characteristic information of the first wafer includes:

[0072] In order to precisely capture the defect abnormalities in the characteristic regions of the first wafer, the surface of the first wafer is divided into multiple local regions.

[0073] After the surface of the first wafer is divided into multiple local regions, multiple local regions are also correspondingly generated on the defect distribution map, and the defect distribution map corresponding to each local region shows the defect distribution in that local region.

[0074] Using a spatial clustering algorithm to analyze the defect distribution map to obtain the local defect density of each local region.

[0075] It should be noted that traditional spatial clustering algorithms (Ordering points to identify the clustering structure, OPTICS) focus on global density clustering. In this embodiment, OPTICS is used to perform density analysis on each local area to finely capture defect anomalies in specific areas, which is particularly applicable to the subtle changes that may occur between different areas in wafer production.

[0076] Specifically, using a density estimation method (such as OPTICS), analyze the defect distribution map corresponding to each local area respectively, and a density-based clustering structure ranking will be obtained. The clustering structure ranking includes multiple ranking points, and each ranking point corresponds to a defect. The density of defects in the local area (i.e., local defect density) is determined by calculating the reachable distance between each ranking point and other ranking points in the local area. This method is applicable to discovering clustering structures with different densities.

[0077] Based on the local defect density of each local area, obtain the first defect density information of the first wafer.

[0078] After calculating the local defect density of each local area in the above manner, record each local area and its corresponding local defect density in the first defect density information of the first wafer.

[0079] Among them, as Figure 2 shown, divide the surface of the first wafer into multiple local areas, including:

[0080] Taking the center of the surface of the first wafer as the center, divide a circular local area in the central area of the surface of the first wafer, that is, as Figure 2 shown, at the center of the first wafer, first divide a circular local area concentric with the first wafer.

[0081] Divide the peripheral area of the circular local area into multiple annular local areas. Among them, the multiple annular local areas are concentrically arranged, and the radius of the circular local area is equal to the annular radius of each annular local area.

[0082] That is, divide the annular area on the first wafer, except for the circular local area, into multiple concentrically arranged annular local areas. The center of each annular local area is the center of the surface of the first wafer, and the annular radius of each annular local area is equal, and the annular radius of the annular local area is the same as the radius of the circular local area.

[0083] It should be noted that the radius of the circular local area and the annular radius of the annular local area can be determined according to the size of the first wafer and the number of annular local areas on the first wafer. For example, asFigure 2 As shown, the circular radii of the circular local area and the annular local area are 3 mm. Figure 2 Among them, r, p, q, etc. all represent sorting points or defects on the wafer.

[0084] In some embodiments, intercepting the first wafer according to the defect feature information and the interception index includes:

[0085] When the first condition is satisfied, it is determined to intercept the first wafer, that is, after calculating the local defect density corresponding to each local area, by determining whether the first condition is satisfied, it is determined whether to intercept the first wafer. When the first condition is satisfied, the first wafer is intercepted. When the first condition is not satisfied, it is determined that the quality of the first wafer is qualified, no interception is performed, and the shipping process continues.

[0086] Wherein, the first condition includes at least one of the following:

[0087] Condition 1: The first defect density information indicates that the local defect density of the local area of the first wafer exceeds the threshold density range.

[0088] Wherein, the threshold density range can be understood as the normal defect density fluctuation range, that is, the normal density threshold range. In Condition 1, using a statistical model, such as the chi-square test method, if it is found that the defect density of any local area of the first wafer exceeds this threshold density range, then it is considered that the quality of the first wafer is unqualified.

[0089] Optionally, different threshold density ranges can be set for different local areas.

[0090] Condition 2: The first defect density information indicates that the density difference between the local defect density of the first local area and the local defect density of the second local area of the first wafer is greater than a preset density difference, wherein the first local area is one of the multiple local areas on the surface of the first wafer, and the second local area is one of the multiple local areas on the surface of the first wafer other than the first local area.

[0091] It can be understood that the first local area and the second local area are any two local areas. In Condition 2, the density difference between any two local areas is determined respectively. If the density difference is too large, greater than the preset density difference, then a highly concentrated area between multiple local areas can be identified. At this time, it is also considered that the quality of the first wafer is unqualified. Please refer to Figure 3 , Figure 3 which is a schematic diagram of the defect distribution on the first wafer. It can be seen that the density difference between the annular local area and the circular local area closest to the edge is too large, and the quality of the first wafer is unqualified.

[0092] In some embodiments, the method further includes:

[0093] Obtaining second defect density information of a second wafer, where the second wafer is a qualified wafer of a previous batch of the first wafer.

[0094] It should be noted that the second wafer may be a qualified wafer of a previous batch of the first wafer, or may be qualified wafers of multiple adjacent batches before the first wafer, that is, wafers not intercepted.

[0095] Among them, the meanings of the second defect density information and the first defect density information are basically the same, and they have the same parameters, except for the specific parameter values.

[0096] Performing regression analysis on the second defect density information to obtain the threshold density range.

[0097] The second defect density information can be understood as historical detection data. Based on this historical detection data, regression analysis is performed on the defect distribution of the second wafer to obtain the range of defect density fluctuations in wafers of normal batches, that is, the threshold density range.

[0098] It should also be noted that in the case where the first wafer is a qualified wafer, the first defect density information can be used as historical detection data to re-determine the threshold density range of the next batch of wafers. That is, in the present embodiment, the threshold density range of defects on the wafer is dynamically adjusted by combining time series and historical detection data of historical batches.

[0099] Since the wafer products of different batches are not uniform in terms of production process, raw materials, etc., the threshold density ranges of their defect distributions are also different. Therefore, it is necessary to adaptively adjust the threshold density range using historical detection data instead of a fixed range. Through continuous data learning and feedback mechanisms, the algorithm can identify hidden high-density clustering regions that are easily overlooked.

[0100] Optionally, in this embodiment, a machine learning algorithm is introduced and trained with historical detection data to obtain the output threshold density range, so that the density deviation threshold can be adaptively adjusted instead of a fixed threshold to cope with changes in different wafer batches, and it can more accurately distinguish random defects from true risk defects.

[0101] In some embodiments, the defect distribution map is analyzed to obtain defect feature information of the first wafer, including:

[0102] Performing feature extraction on the defect distribution map to obtain first cluster defect information of the first wafer;

[0103] Among them, the first cluster defect information includes at least one of the following:

[0104] The morphological characteristics of cluster defects on the first wafer;

[0105] The scale quantity of cluster defects on the first wafer;

[0106] The distribution size of cluster defects on the first wafer;

[0107] The distribution density of cluster defects on the first wafer;

[0108] Among them, the morphological characteristics of cluster defects include at least one of the following:

[0109] The cluster defects exhibit linear characteristics;

[0110] The cluster defects exhibit annular characteristics;

[0111] The cluster defects exhibit local cluster-like characteristics.

[0112] That is, the defects of abnormal clusters usually show a highly concentrated linear, annular or high-density distribution in local areas. Compared with randomly distributed defects, their morphological characteristics are more obvious.

[0113] Exemplarily, please refer to Figure 4 , Figure 4 which is a schematic diagram of the defect distribution on the first wafer. It can be seen from Figure 4 that obvious local cluster-like characteristics appear on the first wafer.

[0114] In some embodiments, intercepting the first wafer according to the defect characteristic information and the interception index includes:

[0115] Under the condition of meeting the second condition, intercepting the first wafer, that is, after obtaining the above-mentioned first cluster defect information, by judging whether the second condition is met, determining whether to intercept the first wafer. Under the condition of meeting the second condition, intercepting the second wafer. Under the condition of not meeting the second condition, determining that the quality of the first wafer is qualified and not intercepting, and continuing the shipping process.

[0116] Among them, the second condition includes at least one of the following:

[0117] Condition 3: The morphological characteristics of cluster defects on the first wafer indicated by the first cluster defect information conform to the morphological characteristics of cluster defects on the third wafer, and the third wafer is a wafer with unqualified quality in the previous batch of the first wafer.

[0118] Among them, the third wafer is a wafer with unqualified quality in one or more batches before the first wafer, that is, the intercepted wafer.

[0119] Under this condition, the morphological features of the cluster defects on the third wafer exhibit linear features, circular features, and local cluster-like features. The morphological features of the cluster defects on the third wafer are used as historical matching data, and the morphological features of the cluster defects on the first wafer indicated by the first cluster defect information of the first wafer are matched with the historical matching data. If they match, it is determined that the quality of the first wafer is unqualified.

[0120] Condition 4: The scale quantity of the cluster defects on the first wafer indicated by the first cluster defect information is greater than the threshold quantity, and the threshold quantity is determined according to the scale quantity of the cluster defects on the fourth wafer. The fourth wafer is a wafer with qualified quality in the previous batch of the first wafer.

[0121] Among them, the fourth wafer is a wafer with qualified quality in one or more batches before the first wafer, that is, a wafer not intercepted.

[0122] It should be noted that the size of the abnormal cluster defects is small, but the quantity is significantly large, or the size and quantity meet the requirements, but the density is abnormally high.

[0123] Therefore, under this condition, a threshold quantity is determined according to the scale quantity of the cluster defects on the fourth wafer. This threshold quantity can be understood as the scale quantity of the cluster defects in a normal batch with qualified quality. If the scale quantity of the cluster defects on the first wafer indicated by the first cluster defect information is greater than the threshold quantity, it is considered that the quality of the first wafer is unqualified.

[0124] Condition 5: The distribution size of the cluster defects on the first wafer indicated by the first cluster defect information is greater than the threshold size, and the threshold size is determined according to the distribution size of the cluster defects on the fourth wafer. The fourth wafer is a wafer with qualified quality in the previous batch of the first wafer.

[0125] Among them, the fourth wafer is a wafer with qualified quality in one or more batches before the first wafer, that is, a wafer not intercepted.

[0126] Therefore, under this condition, a threshold size is determined according to the distribution size of the cluster defects on the fourth wafer. This threshold size can be understood as the threshold size of the cluster defects in a normal batch with qualified quality. If the distribution size of the cluster defects on the first wafer indicated by the first cluster defect information is greater than the threshold size, that is, the area of the cluster defects exceeds a certain threshold area or the area ratio on the wafer exceeds a certain ratio, it is considered that the quality of the first wafer is unqualified.

[0127] Condition 6: The distribution density of cluster defects on the first wafer indicated by the first cluster defect information is greater than the threshold density, where the threshold density is determined according to the distribution density of cluster defects on the fourth wafer, and the fourth wafer is a qualified wafer of a previous batch of the first wafer.

[0128] Among them, the fourth wafer is a qualified wafer in one or more batches before the first wafer, that is, a wafer not intercepted.

[0129] Therefore, in this condition, a threshold density is determined according to the distribution density of cluster defects on the fourth wafer. This threshold density can be understood as the threshold density of cluster defects in a normal qualified batch. If the distribution density of cluster defects on the first wafer indicated by the first cluster defect information is greater than the threshold density, for example, the distribution density of cluster defects on the first wafer is 2.5 times or 3 times the threshold density, it is considered that the first wafer is unqualified in quality.

[0130] That is, in the second condition, the first cluster defect information of the first wafer is compared with historical data. If the corresponding interception index is met, the first wafer is marked as "NG", that is, unqualified in quality, and the first wafer or the batch where the first wafer is located is intercepted.

[0131] In this embodiment, when judging the first cluster defect information, a multi-dimensional analysis method is adopted, and comprehensive judgment is carried out by combining factors such as the morphological characteristics, scale quantity, distribution size, and distribution density of the density. Moreover, the above thresholds can be flexibly adjusted according to the production batch to adapt to the quality control standards of different batches.

[0132] In summary, the embodiment of the present invention constructs an automated wafer shipment interception method, which can analyze the defect distribution data of batch wafers in real time, and perform interceptions with different levels of strictness according to batches with different shipment requirements. When an abnormal aggregation pattern is found, it automatically triggers a Hold and generates a detailed aggregation analysis report for engineers to review. By analyzing the wafer quality through local defect density and / or cluster defect characteristics to determine whether it exceeds the control standard, and then performing interception, compared with the method of intercepting only based on the overall defect density, it can more effectively identify and intercept wafers. By continuously optimizing the density threshold through historical data, the capture accuracy of risk defects is improved. Through local comparative analysis of the aggregation density of each region on the wafer surface, the risk of missing local aggregation defects due to uniform particle distribution across the entire wafer can be effectively avoided. Combining historical batch data with real-time defect distribution analysis, an adaptive interception rule is constructed, reducing the subjectivity and inconsistency of manual judgment, and can perform interceptions with different levels of strictness according to batches with different shipment requirements, improving the efficiency and accuracy of shipment interception. It helps to improve the limitations in existing wafer shipment interception, especially when dealing with complex cluster-like aggregation defects, and can provide more refined control and judgment criteria.

[0133] As Figure 5 described above, the embodiment of the present invention further provides a wafer defect detection device, including:

[0134] A first processing module 501, configured to scan a first wafer to obtain a defect distribution map of the wafer;

[0135] A second processing module 502, configured to analyze the defect distribution map to obtain defect characteristic information of the first wafer, where the defect characteristic information includes at least one of the following: first defect density information, first cluster defect information;

[0136] A third processing module 503, configured to intercept the first wafer according to the defect characteristic information and an interception index.

[0137] Optionally, the second processing module 502 includes:

[0138] A first processing unit, configured to divide the surface of the first wafer into multiple local regions;

[0139] A second processing unit, configured to analyze the defect distribution map by using a spatial clustering algorithm to obtain the local defect density of each local region;

[0140] A third processing unit, configured to obtain the first defect density information of the first wafer according to the local defect density of each local region.

[0141] Optionally, the first processing unit is specifically configured to:

[0142] Taking the center of the surface of the first wafer as the center, a circular local area is divided in the central area of the surface of the first wafer;

[0143] The peripheral area of the circular local area is divided into a plurality of annular local areas;

[0144] Wherein, the plurality of annular local areas are concentrically arranged, and the radius of the circular local area is equal to the annular radius of each of the annular local areas.

[0145] Optionally, the third processing module 503 includes:

[0146] A fourth processing unit, configured to determine to intercept the first wafer when a first condition is met;

[0147] Wherein, the first condition includes at least one of the following:

[0148] The first defect density information indicates that the local defect density of a local area of the first wafer exceeds a threshold density range;

[0149] The first defect density information indicates that the density difference between the local defect density of a first local area and the local defect density of a second local area of the first wafer is greater than a preset density difference, wherein the first local area is one of a plurality of local areas on the surface of the first wafer, and the second local area is one of the plurality of local areas on the surface of the first wafer other than the first local area.

[0150] Optionally, the device further includes:

[0151] A first acquisition module, configured to acquire second defect density information of a second wafer;

[0152] A fourth processing module, configured to perform regression analysis on the second defect density information to obtain the threshold density range;

[0153] Wherein, the second wafer is a wafer with qualified quality in a previous batch of the first wafer.

[0154] Optionally, the second processing module 502 includes:

[0155] A fifth processing unit, configured to perform feature extraction on the defect distribution map to obtain first cluster defect information of the first wafer;

[0156] Wherein, the first cluster defect information includes at least one of the following:

[0157] The morphological characteristics of cluster defects on the first wafer;

[0158] The scale quantity of cluster defects on the first wafer;

[0159] The distribution size of cluster defects on the first wafer;

[0160] The distribution density of cluster defects on the first wafer.

[0161] Optionally, the third processing module 503 includes:

[0162] A sixth processing unit, configured to intercept the first wafer when a second condition is met;

[0163] Wherein, the second condition includes at least one of the following:

[0164] The morphological characteristics of the cluster defects on the first wafer indicated by the first cluster defect information conform to the morphological characteristics of the cluster defects on the third wafer, where the third wafer is a wafer with unqualified quality in a previous batch of the first wafer;

[0165] The scale quantity of the cluster defects on the first wafer indicated by the first cluster defect information is greater than a threshold quantity, and the threshold quantity is determined according to the scale quantity of the cluster defects on the fourth wafer;

[0166] The distribution size of the cluster defects on the first wafer indicated by the first cluster defect information is greater than a threshold size, and the threshold size is determined according to the distribution size of the cluster defects on the fourth wafer;

[0167] The distribution density of the cluster defects on the first wafer indicated by the first cluster defect information is greater than a threshold density, and the threshold density is determined according to the distribution density of the cluster defects on the fourth wafer;

[0168] Wherein, the fourth wafer is a wafer with qualified quality in a previous batch of the first wafer.

[0169] Optionally, the morphological characteristics of the cluster defects include at least one of the following:

[0170] The cluster defects exhibit a linear feature;

[0171] The cluster defects exhibit an annular feature;

[0172] The cluster defects exhibit a local cluster-like feature.

[0173] It should be noted that the wafer defect detection device provided in the embodiments of the present invention is a device capable of executing the above-mentioned wafer defect detection method. Therefore, all embodiments of the above-mentioned wafer defect detection method are applicable to this device and can achieve the same or similar technical effects.

[0174] Such asFigure 6 As shown in the figure, an embodiment of the present invention further provides a wafer defect detection device, including: a processor 601; and a memory 603 connected to the processor 601 through a bus interface 602, where the memory 603 is used to store programs and data used by the processor 601 when performing operations, and the processor 601 calls and executes the programs and data stored in the memory 603.

[0175] Wherein, the transceiver 604 is connected to the bus interface 602 and is used to receive and send data under the control of the processor 601. Specifically, the processor 601 is used to read the programs in the memory 603, and the processor 601 is used to perform the following processes:

[0176] Scan the first wafer to obtain a defect distribution map of the first wafer;

[0177] Analyze the defect distribution map to obtain defect feature information of the first wafer, where the defect feature information includes at least one of the following: first defect density information, first cluster defect information;

[0178] Intercept the first wafer according to the defect feature information and the interception index.

[0179] Optionally, the processor 601 is used to:

[0180] Divide the surface of the first wafer into multiple local areas;

[0181] Use a spatial clustering algorithm to analyze the defect distribution map to obtain the local defect density of each local area;

[0182] Obtain the first defect density information of the first wafer according to the local defect density of each local area.

[0183] Optionally, the processor 601 is specifically used to:

[0184] Take the center of the surface of the first wafer as the center and divide a circular local area in the central area of the surface of the first wafer;

[0185] Divide the peripheral area of the circular local area into multiple annular local areas;

[0186] Wherein, the multiple annular local areas are concentrically arranged, and the radius of the circular local area is equal to the annular radius of each annular local area.

[0187] Optionally, the processor 601 is used to:

[0188] When the first condition is met, determine to intercept the first wafer;

[0189] Wherein, the first condition includes at least one of the following:

[0190] The first defect density information indicates that the local defect density in a local area of the first wafer exceeds the threshold density range;

[0191] The first defect density information indicates that the density difference between the local defect density in a first local area and the local defect density in a second local area of the first wafer is greater than a preset density difference, wherein the first local area is one of multiple local areas on the surface of the first wafer, and the second local area is one of the multiple local areas on the surface of the first wafer other than the first local area.

[0192] Optionally, the processor 601 is further configured to:

[0193] Obtain the second defect density information of a second wafer;

[0194] Perform regression analysis on the second defect density information to obtain the threshold density range;

[0195] Wherein, the second wafer is a wafer with qualified quality in a previous batch of the first wafer.

[0196] Optionally, the processor 601 is configured to:

[0197] Extract features from the defect distribution map to obtain the first cluster defect information of the first wafer;

[0198] Wherein, the first cluster defect information includes at least one of the following:

[0199] The morphological characteristics of cluster defects on the first wafer;

[0200] The scale quantity of cluster defects on the first wafer;

[0201] The distribution size of cluster defects on the first wafer;

[0202] The distribution density of cluster defects on the first wafer.

[0203] Optionally, the processor 601 is configured to:

[0204] When the second condition is met, intercept the first wafer;

[0205] Wherein, the second condition includes at least one of the following:

[0206] The morphological characteristics of the cluster defects on the first wafer indicated by the first cluster defect information conform to the morphological characteristics of the cluster defects on the third wafer, where the third wafer is a wafer with unqualified quality in a previous batch of the first wafer;

[0207] The scale quantity of the cluster defects on the first wafer indicated by the first cluster defect information is greater than the threshold quantity, and the threshold quantity is determined according to the scale quantity of the cluster defects on the fourth wafer;

[0208] The distribution size of the cluster defects on the first wafer indicated by the first cluster defect information is greater than the threshold size, and the threshold size is determined according to the distribution size of the cluster defects on the fourth wafer;

[0209] The distribution density of the cluster defects on the first wafer indicated by the first cluster defect information is greater than the threshold density, and the threshold density is determined according to the distribution density of the cluster defects on the fourth wafer;

[0210] Wherein, the fourth wafer is a wafer with qualified quality in a previous batch of the first wafer.

[0211] Optionally, the morphological characteristics of the cluster defects include at least one of the following:

[0212] The cluster defects exhibit linear characteristics;

[0213] The cluster defects exhibit annular characteristics;

[0214] The cluster defects exhibit local cluster-like characteristics.

[0215] Wherein, in Figure 6 The bus architecture can include any number of interconnected buses and bridges, specifically various circuits represented by one or more processors represented by processor 601 and a memory represented by memory 603 are linked together. The bus architecture can also link together various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art, and thus will not be further described herein. The bus interface provides user interface 605. The transceiver 604 can be multiple elements, that is, including a transmitter and a receiver, and provides a unit for communicating with various other devices on the transmission medium. Processor 601 is responsible for managing the bus architecture and general processing, and memory 603 can store data used by processor 601 when performing operations.

[0216] The above is the preferred embodiment of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle described in the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention.

Claims

1. A wafer defect detection method, characterized in that: include: Scanning a first wafer to obtain a defect distribution map of the first wafer; Analyze the defect distribution map to obtain defect feature information of the first wafer, wherein the defect feature information includes at least one of the following: first defect density information and first cluster defect information; The first wafer is intercepted according to the defect feature information and the interception index.

2. The method according to claim 1, characterized in that Analyzing the defect distribution map to obtain defect feature information of the first wafer includes: dividing the surface of the first wafer into a plurality of local areas; Analyzing the defect distribution map using a spatial clustering algorithm to obtain a local defect density of each local area; First defect density information of the first wafer is obtained according to the local defect density of each of the local areas.

3. The method according to claim 2, characterized in that Dividing the surface of the first wafer into a plurality of local areas, including: Taking the center of the surface of the first wafer as the center of the circle, dividing a circular local area in the central area of ​​the surface of the first wafer; dividing the outer area of ​​the circular local area into a plurality of annular local areas; The plurality of annular local areas are concentrically arranged, and the radius of the circular local area is equal to the annular radius of each of the annular local areas.

4. The method according to claim 1, characterized in that: Intercepting the first wafer according to the defect feature information and the interception index includes: When the first condition is met, determining to intercept the first wafer; The first condition includes at least one of the following: The first defect density information indicates that a local defect density of a local area of ​​the first wafer exceeds a threshold density range; The first defect density information indicates that a density difference between a local defect density of a first local area and a local defect density of a second local area of ​​the first wafer is greater than a preset density difference, wherein the first local area is one of multiple local areas on the surface of the first wafer, and the second local area is one of multiple local areas on the surface of the first wafer other than the first local area.

5. The method according to claim 4, characterized in that The method further comprises: Acquiring second defect density information of a second wafer; Performing regression analysis on the second defect density information to obtain the threshold density range; The second wafer is a wafer of qualified quality from a previous batch of the first wafer.

6. The method according to claim 1, characterized in that Analyzing the defect distribution map to obtain defect feature information of the first wafer includes: Performing feature extraction on the defect distribution map to obtain first cluster defect information of the first wafer; The first cluster defect information includes at least one of the following: morphological characteristics of cluster defects on the first wafer; the scale and quantity of cluster defects on the first wafer; a distribution size of cluster defects on the first wafer; The distribution density of cluster defects on the first wafer.

7. The method according to claim 1, characterized in that Intercepting the first wafer according to the defect feature information and the interception index includes: When the second condition is met, intercepting the first wafer; The second condition includes at least one of the following: The morphological features of the cluster defects on the first wafer indicated by the first cluster defect information are consistent with the morphological features of the cluster defects on the third wafer, wherein the third wafer is a wafer of unqualified quality in a previous batch of the first wafer; The number of cluster defects on the first wafer indicated by the first cluster defect information is greater than a threshold number, and the threshold number is determined according to the number of cluster defects on the fourth wafer; The distribution size of the cluster defects on the first wafer indicated by the first cluster defect information is greater than a threshold size, and the threshold size is determined according to the distribution size of the cluster defects on the fourth wafer; The distribution density of cluster defects on the first wafer indicated by the first cluster defect information is greater than a threshold density, and the threshold density is determined according to the distribution density of cluster defects on a fourth wafer; The fourth wafer is a wafer of qualified quality from a previous batch of the first wafer.

8. The method according to claim 7, characterized in that The morphological characteristics of cluster defects include at least one of the following: The cluster defects are linear in nature; The cluster defect has a ring-shaped feature; The cluster defects are characterized by local clusters.

9. A wafer defect detection device, characterized in that: include: A first processing module is used to scan a first wafer to obtain a defect distribution map of the wafer; A second processing module is used to analyze the defect distribution map to obtain defect feature information of the first wafer, wherein the defect feature information includes at least one of the following: first defect density information and first cluster defect information; The third processing module is used to intercept the first wafer according to the defect feature information and the interception index.

10. A wafer defect detection device, characterized in that: include: 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 in the wafer defect detection method as claimed in any one of claims 1 to 8.

Images