Method, device and computer program product for wafer surface defect detection

Abstract

The technical scheme of the present application provides a wafer surface defect detection method, device and computer program product, wherein the wafer surface defect detection method comprises: acquiring a surface image of a test wafer under standard image acquisition conditions, wherein the test wafer surface has a plurality of periodic patterns; comparing the standard template image with the minimum period images in the surface image of the test wafer one by one; and judging defects according to the analysis comparison result: when the difference between the standard template image and the minimum period images in the surface image of the test wafer exceeds the judgment threshold, judging that the region where the minimum period images in the surface image of the test wafer are located is a defect region. The wafer surface defect detection method provided by the present application improves the defect detection efficiency and improves the product yield.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of semiconductor equipment, in particular to a wafer surface defect detection method and device and a computer program product thereof. BACKGROUND

[0002] In the process of integrated circuit manufacturing, the glue coating and developing equipment is the equipment in the photoetch process, including glue coating machine, glue spraying machine and developing machine, etc., which acts on the input and output links of photoetching, specifically including photoresist coating by glue coating machine before exposure, and developing by developing machine after exposure. These are all online operations of glue coating and developing equipment and photoetching machine. In the process of wafer manufacturing, each process may cause pollution to the wafer.

[0003] For example, the defects introduced on the semiconductor substrate by glue coating include particle, crystallization, bubble and other particle type defects introduced by the bottom antireflection layer and photoresist itself.

[0004] In addition, radial defects are generated during glue coating due to the particle foreign matter on the wafer surface. The introduced defects on the wafer surface generally cause the wafer surface to protrude, causing glue coating (including photoresist and antireflection layer) to be poor. Such defects are generally particles. Such particles can be brought from the previous layer process. For example, particles or flakes falling from the inner surface of the cavity in dry etching (plasma etching), or particles brought by physical vapor deposition (PVD) process, etc. In the process of glue coating, the rotation speed, the horizontal state of the wafer, the amount of glue coating, etc. will also cause the wafer surface defects to be generated.

[0005] The defects introduced on the semiconductor substrate by the developing process mainly include the following two types: (1) water mist (developer mist) generated due to back splash of the developing liquid, causing a small area of overdeveloped edge region of the wafer, and (2) photoresist residues left on the wafer surface without being washed away by the developing washing.

[0006] In addition, there are peeling defects caused by poor adhesion to the substrate.

[0007] (1) Poor development of photoresist, because the development or development washing is not good, and the dissolved and partially dissolved photoresist residues cannot be effectively taken away from the wafer surface. Usually such defects occur first at the edge of the wafer.

[0008] (2) The lithography process does not have enough depth of focus (DOF), so that when there is a little fluctuation on the wafer surface, hole loss occurs in a certain area.

[0009] (3) The lithography process has a high mask error factor (MEF), generally > > 4.0, when the line width error on the mask reaches a certain degree, hole loss occurs.

[0010] The defects generated in the process of coating and developing have a direct impact on the effect of lithography, thereby having a great influence on the final yield of the entire process. Therefore, defect detection of the coating and developing process plays a crucial role in the entire process. The efficiency of defect detection also has a very direct impact on the efficiency of the entire chip processing process. SUMMARY

[0011] The purpose of the present application is to provide a wafer surface defect detection method and device, which can quickly and efficiently detect the defects on the wafer surface in the chip processing process.

[0012] In order to solve the above technical problems, the present application provides a wafer surface defect detection method, comprising:

[0013] Obtaining the surface image of the test wafer under standard image acquisition conditions, wherein the surface of the test wafer has a plurality of periodic regular patterns;

[0014] Comparing the standard template image with the smallest period image in the surface image of the test wafer one by one;

[0015] Judging the defect according to the analysis comparison result:

[0016] When the difference between the standard template image and the smallest period image in the surface image of the test wafer exceeds the judgment threshold, the area where the smallest period image in the surface image of the test wafer is compared is judged as a defect area.

[0017] Preferably, comprising:

[0018] The standard template image is formed by processing the standard image of the standard wafer, wherein the standard wafer is a non-polluted wafer, and the surface thereof has a plurality of periodic regular patterns similar to the surface of the test wafer;

[0019] The size of the standard template image is related to the smallest period size of the periodic regular pattern.

[0020] Preferably, the standard image of the standard wafer is the surface image of the standard wafer collected under standard image acquisition conditions.

[0021] Preferably, the step of obtaining the standard image acquisition condition comprises:

[0022]

[0023]

[0024]

[0025] Preferably, the step of forming the standard template image comprises:

[0026]

[0027]

[0028]

[0029]

[0030] Preferably, the step of cutting out a periodic regular pattern of a minimum period from the standard image of the standard wafer comprises:

[0031]

[0032] Preferably, after the step of obtaining the surface image of the test wafer under the standard image acquisition condition, and before the step of comparing the standard template image with the minimum period image in the surface image of the test wafer one by one, the method further comprises: performing preliminary positioning on the surface image of the test wafer.

[0033] Preferably, the step of performing preliminary positioning on the surface image of the test wafer comprises:

[0034]

[0035]

[0036] If not, the surface image of the test wafer is rotated by a certain angle.​​​​​​​​​​

[0037] Preferably, the step of comparing the standard template image with the minimum period images in the surface image of the test wafer one by one comprises:

[0038] The minimum period images in the surface image of the test wafer are shape-matched one by one with the standard template image to select minimum period images to be detected;

[0039] The gray scale values of the selected minimum period images to be detected are differentially compared with the gray scale values of the standard template image to obtain the difference values.

[0040] Preferably, the different image acquisition conditions can include different factors settings of bright-dark field type, LED brightness, camera exposure parameter or / and camera gain.

[0041] Preferably, the judging defects comprises judging dark defects and bright defects.

[0042] In another aspect, the technical scheme of the present application further provides a wafer surface defect detection device, comprising:

[0043] An image storage unit adapted to store a standard template image;

[0044] An image acquisition unit adapted to acquire a surface image of a test wafer under standard image acquisition conditions, wherein the test wafer surface has a plurality of periodic regular patterns;

[0045] A test comparison unit adapted to compare the standard template image with the minimum period images in the surface image of the test wafer one by one;

[0046] A defect judging unit adapted to judge defects according to the comparison results of the comparison module, wherein,

[0047] When the difference between the standard template image and the minimum period images in the surface image of the test wafer exceeds a judging threshold, the region where the compared minimum period images in the surface image of the test wafer are located is judged as a defect region.

[0048] Preferably, the image acquisition unit is further adapted to acquire surface images of a standard wafer under different image acquisition conditions, wherein the standard wafer is a non-contaminated wafer, and the surface thereof has a plurality of periodic regular patterns similar to the surface of the test wafer;

[0049] The size of the standard template image is related to the minimum period size of the periodic regular patterns.

[0050] Preferably, it further comprises a standard model image conversion unit, comprising:

[0051] a basic image storage unit adapted to store images of standard wafers respectively collected under different image collection conditions;

[0052] a gray value calculation unit adapted to calculate an average gray value of each image of the standard wafers;

[0053] a standard image matching unit adapted to compare the difference between the average gray value of each image and a preset standard gray value, and select a surface image of a standard wafer with an average gray value close to the standard gray value as a standard image of the standard wafer.

[0054] Preferably, the method further comprises: a standard image collection condition setting module adapted to set the environmental condition corresponding to the image of the standard wafer with the average gray value close to the standard value as a standard image collection condition.

[0055] Preferably, the standard model image processing unit further comprises:

[0056] an initial template image extraction unit adapted to extract a periodic regular pattern of a minimum period from the standard image as an initial template image;

[0057] a pre-template image extraction unit adapted to extract n periodic regular patterns of minimum periods from the standard image as pre-template images based on a preset matching score of the initial template image and a pre-template image size;

[0058] a matching value screening unit adapted to eliminate the pre-template images with the highest and lowest matching scores with the initial template image;

[0059] a best image selection unit adapted to perform image superposition processing on the remaining n-2 pre-template images, obtain an average image and a standard deviation image, and finally combine to obtain a best image as a selected periodic regular pattern of a minimum period.

[0060] Preferably, the minimum periodic regular pattern calculation unit is adapted to: based on the pattern regularity of the wafer image, use Fourier transform to obtain conjugate coefficients, transform the image information from a spatial domain to a frequency domain, utilize a two-dimensional sinusoidal wave to fit the fluctuation of the gray value, obtain the periodic regular image size in the horizontal and vertical directions, and calculate the periodic regular pattern size.

[0061] Preferably, the method further comprises a surface image preliminary positioning unit of the test wafer adapted to search for a position area of a wafer positioning groove in the surface image of the test wafer, and judge whether the position of the wafer positioning groove is accurate;

[0062] If not, the surface image of the test wafer is rotated by a certain angle.

[0063] Preferably, the test comparison module comprises:

[0064] The shape matching unit is adapted to perform shape matching by using the periodic regular pattern with the minimum period in the standard image, and select the position of the periodic regular pattern with the minimum period in the surface image of the test wafer that needs to be detected;

[0065] The difference unit is adapted to perform difference between the periodic regular pattern with the minimum period in the surface image of the test wafer that needs to be detected and the average image.

[0066] In still another aspect, the technical scheme of the present application further provides a device for wafer surface defect detection, comprising:

[0067] a processor,

[0068] a memory for storing a computer program or instructions, which, when executed by the processor, implement the steps of the method according to the above to perform defect detection on the object.

[0069] The technical scheme of the present application further provides a computer program product stored on a computer-readable storage medium and comprising a computer program or instructions, which, when executed by the processor, implement the steps of the method according to the above.

[0070] Compared with the prior art, the method for wafer surface defect detection provided by the technical scheme of the present application has the following beneficial effects:

[0071] The standard template image is formed as a standard comparison object for test wafer surface defect detection, and the threshold for judging defects is preset, so that the detection process can be automated, a fast detection standard creation process is realized, and the degree of automation is fast and efficient, thereby improving the efficiency of defect detection and improving product yield;

[0072] Further, by comparing the minimum period images in the surface image of the test wafer one by one to detect and judge defects, the defect area and the chip with defects can be accurately detected, and the detection accuracy is improved;

[0073] Further, the standard template image is obtained under standard acquisition conditions, has good representativeness of a wafer surface image without defects, and serves as a basis for judging whether a defect wafer surface has defects, thereby ensuring the accuracy of detection;

[0074] Further, the standard acquisition conditions are judged by the average gray value of the surface images of multiple standard wafers acquired under different image acquisition conditions, further ensuring the completeness of the standard template image without defects, and further ensuring the accuracy of detection;

[0075] Furthermore, the standard template image is obtained by processing the preceding template image of a periodic pattern with n minimum periods, and the preceding templates with the highest and lowest matching scores are removed. It is synthesized by the average image and the standard variance image, which further ensures the representativeness of the standard template image and guarantees the accuracy of detection.

[0076] Furthermore, by obtaining the minimum periodic pattern of the wafer surface image through Fourier transform, the minimum periodic pattern of the wafer surface can be quickly and accurately identified, enabling the inspection equipment to automatically and quickly identify the chip image in the wafer surface image.

[0077] Furthermore, before formal testing, the location of the notch in the surface image of the test wafer can be identified to determine whether the position of the test wafer is accurate. The initial positioning step is integrated into the process of detection by acquiring images, ensuring that all steps of the defect detection process can be uniformly integrated into the automated program to complete fast and efficient detection. Attached Figure Description

[0078] Appendix Figure 1 This is a schematic diagram of the surface of a standard wafer and / or a test wafer provided in an embodiment of the present invention.

[0079] Appendix Figure 2 This is a schematic flowchart of a method for detecting defects on a wafer surface provided in an embodiment of the present invention.

[0080] Appendix Figure 3 This is a schematic flowchart illustrating the method for forming a standard template image provided in an embodiment of the present invention.

[0081] Appendix Figure 4 This is a schematic diagram illustrating the process of obtaining standard image acquisition conditions provided in an embodiment of the present invention.

[0082] Appendix Figure 5 This is a schematic flowchart illustrating the method for forming the standard template image provided in an embodiment of the present invention.

[0083] Appendix Figure 6 This is a schematic diagram of a wafer surface defect detection device provided in an embodiment of the present invention. Detailed Implementation

[0084] Numerous specific details are set forth in the following description to provide a full understanding of the invention. However, the invention can be practiced in many other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0085] Secondly, the present application is described in detail by using a schematic diagram, and in the detailed description of the embodiments of the present application, the schematic diagram is only an example, which should not limit the protection scope of the present application.

[0086] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the specific embodiments of the present application are described in detail below with reference to the drawings.

[0087] In the present application, the method of template matching is mainly used to realize the automatic measurement of the defects on the wafer surface. Template matching refers to finding the similar part on the tested image to the template image through the comparison between the template image and the tested image, which is realized by calculating the similarity of the target in the template image and the tested image, and the predefined target can be quickly defined in the tested image. The main idea of matching is to use a target prototype, create a template according to it, search for the most similar target in the tested image, and find the area closest to the mean or variance of the template. The specific method of template matching also needs to be determined according to the characteristics of the application scene, the specific image data and the matching task requirements.

[0088] Specifically, in combination with Figures 1 to 5 In the present embodiment, a wafer surface defect detection method is provided. The wafer surface defect detection method is applied to an upper computer with image analysis function and unified deployment and sending operation instruction, a device for collecting wafer images, and a wafer detection system of a wafer carrying table. The wafer detection system can be a separate wafer defect detection device, or a system or device integrated with wafer measurement function or other image analysis function. Those skilled in the art can make possible equivalent changes or simple modifications under the condition of understanding the essence of the technical solution.

[0089] Specifically, in the present embodiment, the device for collecting wafer images is a line scan camera. The wafer image collection work in the present embodiment cooperates with the device providing auxiliary functions, and the device providing auxiliary functions further includes a wafer carrying table cooperating with the line scan camera.

[0090] The wafer to be collected is stably placed on the wafer carrying table. The wafer carrying table is also adapted to cooperate with a linear motion device, and can provide the wafer with stable linear motion of fixed step length, so as to cooperate with the line scan camera to scan the wafer surface row by row or column by column.

[0091] Referring to Figure 1Fig. 1 shows a schematic diagram of a wafer surface provided by the detection of the present embodiment. In the present embodiment, it is a wafer after the glue coating and developing step. Those skilled in the art can understand that the wafer 90 after the glue coating and developing step has a plurality of chip patterns 44 arranged in a periodic pattern. In the present embodiment, the chip patterns 44 are approximately rectangular, and the size is the same. The technical principle of the present concept is described in this embodiment.

[0092] Those skilled in the art can deduce how the technical solution of the present application applies to wafers with other types of chip patterns without departing from the content of the present application and on the basis of the technical essence of the present application.

[0093] Further, in the present embodiment, the wafer 90 includes two types of wafers:

[0094] Standard wafer: The standard wafer is a wafer that has undergone a glue coating and developing step and has been determined to be free of contamination and defects through detection.

[0095] Test wafer: The test wafer is a wafer that has undergone a glue coating and developing step and needs to be detected for surface defects during the processing process. The test wafer can have defects formed during the glue coating and developing step. If the defect detection is qualified, the test wafer needs to be processed as soon as possible. If too many defects are found during defect detection, the test wafer is prevented from entering the next step of the process.

[0096] Figure 2 Fig. 2 shows a flowchart of a wafer surface defect detection method provided by the present embodiment. The wafer surface defect detection method includes:

[0097] Step F100: Obtain a surface image of a test wafer under standard image acquisition conditions, wherein the surface of the test wafer has a plurality of periodic patterns;

[0098] Step F200: Compare the standard template image with the smallest periodic image in the surface image of the test wafer one by one;

[0099] Step F300: Determine defects based on the analysis of the comparison results:

[0100] When the difference between the standard template image and the smallest periodic image in the surface image of the test wafer exceeds the judgment threshold, the region where the smallest periodic image in the surface image of the test wafer is located is determined to be a defect region.

[0101] The following reference Figures 1 to 2 , reference Figures 3 to 5 Fig. 1 shows a schematic diagram of a wafer surface provided by the detection of the present embodiment. In the present embodiment, it is a wafer after the glue coating and developing step. Those skilled in the art can understand that the wafer 90 after the glue coating and developing step has a plurality of chip patterns 44 arranged in a periodic pattern. In the present embodiment, the chip patterns 44 are approximately rectangular, and the size is the same. The technical principle of the present concept is described in this embodiment.

[0102] Specifically, referring to Figure 3 , Figure 3 is a flowchart of the method for forming a standard template image provided in the present embodiment. Before performing the steps of wafer surface defect detection shown in Figure 2 , the standard template image in step F200 is formed first. Specifically, the method for forming a standard template image includes:

[0103] Step B10: providing a standard wafer and obtaining a standard image of the standard wafer under standard image acquisition conditions;

[0104] Step B20: processing the standard image to form a standard template image.

[0105] The method for forming a standard template image is described in detail as follows.

[0106] Step B10 is performed: obtaining a standard image of a standard wafer under standard image acquisition conditions;

[0107] In combination with Figure 2 , referring to Figure 4 , the step of obtaining standard image acquisition conditions in step B10 in the present embodiment includes:

[0108] Step S101: acquiring images of a standard wafer under different image acquisition conditions respectively;

[0109] Step S102: calculating the average gray value of each image;

[0110] Step S103: taking the environmental conditions corresponding to the image with an average gray value close to a standard value as the standard image acquisition conditions.

[0111] Specifically, step S101 is performed: acquiring images of a standard wafer under different image acquisition conditions respectively;

[0112] In the present step, the standard wafer is a wafer that has undergone a gumming and developing step and has been determined to be free of contamination and defects through detection.

[0113] Specifically, the present step includes setting multiple different acquisition environmental conditions. Specifically, multiple sets of acquisition environmental conditions can be added through an upper computer application program, and the multiple sets of acquisition environmental conditions can include the setting of different bright field and dark field types, LED brightness, camera exposure parameters, camera gain, and the like.

[0114] In the present embodiment, the multiple different acquisition environmental conditions are multiple different brightness environments. The multiple different brightness environments can be automatically filled with multiple combinations with LED brightness as a fixed step. Further, in the present embodiment, 7-8 different brightness environments are specifically set for the same standard wafer as the environment for acquiring images of the standard wafer.

[0115] Then, in this step, the host computer sends an image acquisition signal instruction, and transports the standard wafer 90 stably placed on the supporting table to the line scan camera for scanning, and drives the linear motion motor to cooperate with the scanning work of the line scan camera. The line scan camera obtains a plurality of complete surface images of the standard wafer under different acquisition environment conditions.

[0116] Step S102 is performed: the average gray value of each image is calculated;

[0117] The gray values of a plurality of points on the images of the standard wafer acquired under different acquisition environment conditions are acquired by image analysis, so as to calculate the average gray value of each image of the standard wafer.

[0118] The plurality of points acquired on the images of the standard wafer are uniformly distributed at various positions of the images of the standard wafer, including the center of the images of the standard wafer, the circumferences of different diameters around the center, different central angles on the circumference, and the edge of the wafer, etc., so as to comprehensively reflect the gray values at various positions of each image of the standard wafer as much as possible.

[0119] The average gray value of each image of the standard wafer acquired under different acquisition environment conditions is calculated by calculating the average gray value. Specifically, the average gray value ranges from 0 to 255.

[0120] In this embodiment, the images of the standard wafer under seven different acquisition environment conditions are acquired, and the average gray values thereof are calculated respectively. For example:

[0121] The average gray value of the image of the standard wafer acquired under the first acquisition environment condition is 126; the average gray value of the image of the standard wafer acquired under the second acquisition environment condition is 150; the average gray value of the image of the standard wafer acquired under the third acquisition environment condition is 127; the average gray value of the image of the standard wafer acquired under the fourth acquisition environment condition is 143; the average gray value of the image of the first standard wafer acquired under the fifth acquisition environment condition is 140; the average gray value of the image of the first standard wafer acquired under the sixth acquisition environment condition is 130; and the average gray value of the image of the first standard wafer acquired under the seventh acquisition environment condition is 146.

[0122] Step S103 is performed: the environment condition corresponding to the image with the average gray value close to the standard value is taken as the standard image acquisition condition.

[0123] In this step, the standard value of the average gray value is set according to the experience value. If the standard value of the average gray value is set too high or too low, the contrast between the defect area and other areas will not be obvious, and the defect will be misjudged. Preferably, the standard value of the average gray value can range from 120 to 136.

[0124] In this embodiment, the standard value for the average grayscale value is set to 128. In other embodiments, the standard value for the average grayscale value can also be set to other values ​​based on experience.

[0125] Furthermore, in this step, the illumination conditions corresponding to an average grayscale value close to 128 are selected as the optical image acquisition conditions. The host computer image analysis system compares the grayscale information of the wafer images acquired under the first to seventh acquisition environment conditions, and selects an average grayscale value close to 128 as the optimal optical image acquisition condition. That is, the third acquisition environment condition with an average grayscale value of 127 for the standard wafer image is the standard image of the standard wafer under the standard image acquisition conditions.

[0126] If the surface image of the standard wafer is acquired under the specified standard image acquisition conditions, then it is the standard image of the standard wafer.

[0127] Next, step B20 is performed: the standard image is processed to form a standard template image.

[0128] In this embodiment, the standard template image is related to the minimum period size of the periodic pattern.

[0129] For details, please refer to Figure 5 As shown, the method for forming the standard template image in this embodiment includes:

[0130] Step S201: Extract a periodic pattern with a minimum period from the standard image of the standard wafer as the initial template image;

[0131] Step S202: Based on the initial template image's preset matching score and the size of the preceding template image, extract the periodic pattern of n minimum cycles from the standard image as the preceding template image;

[0132] Step S203: Remove the preceding template graphs with the highest and lowest matching scores to the initial template graph;

[0133] Step S204: Overlay the remaining n-2 preceding template images to obtain an average image and a standard variance image, and finally merge them to obtain the best image as the standard template image.

[0134] Specifically, step S201 involves extracting a periodic pattern with a minimum period from the standard image of the standard wafer as an initial template image.

[0135] In this embodiment, an image of a standard wafer under optimal optical image acquisition conditions is used as the standard image. Based on the pattern of the pattern, the size of the smallest periodic pattern is determined by an image analysis system, and the periodic pattern is extracted as the pre-template image.

[0136] Specifically, in this embodiment, the image information of the standard image is imported through the host computer application. Based on the image analysis system of the host computer, the conjugate coefficients are obtained by using Fourier transform, and the image information of the standard image is transformed from the spatial domain to the frequency domain. The fluctuation changes of gray values ​​are fitted by a two-dimensional sine wave to obtain the size of the periodic image in the horizontal and vertical directions, thereby calculating the size of the periodic pattern. The pattern with periodicity is the pre-template image, and the size of the periodic pattern is the size of the chip on the wafer surface, including the width and length of the chip.

[0137] Execution step S202: Based on the preset matching score of the initial template image and the size of the preceding template image, extract the periodic pattern of n minimum periods from the standard image as the preceding template image;

[0138] Specifically, in step S202, based on the width and length of the chip obtained in step 201, a complete minimum periodic pattern is extracted from any position of the standard image by the host computer application, that is, an image of a chip is used as the front template image; in this embodiment, in this step, the matching score of the following matching process is preset to 0.4.

[0139] Furthermore, in this embodiment, the position where a complete minimum periodic pattern is selectively extracted during the above process is the center position or center position of the image near the standard wafer.

[0140] Then, extract n complete minimum periodic patterns from other locations in the standard image as matching template images;

[0141] Next, step S203 is performed: remove the preceding template images with the highest and lowest matching scores to the initial template image;

[0142] Specifically, the n matching template images are compared with the preceding template image, and the images with the highest and lowest matching scores are removed, leaving n-2 matching template images;

[0143] Next, step S204 is executed: then the remaining n-2 matching images are overlaid to obtain the average map and the standard deviation map;

[0144] Finally, the mean plot and the standard variance plot are merged to obtain the target template plot, which is the standard template image (GoldenImage) for comparison in the subsequent defect detection steps.

[0145] Next, you can follow Figure 2 The steps of the defect detection method are as follows: call the above standard template image to perform defect detection on the wafer to be tested.

[0146] Further, in the present embodiment, before acquiring the surface image of the test wafer under the standard image acquisition condition, the surface image of the test wafer is further subjected to preliminary positioning, and the step of preliminary positioning the surface image of the test wafer comprises:

[0147] searching for a position area of the wafer positioning groove in the surface image of the test wafer, judging whether the position of the wafer positioning groove is accurate, and if not, rotating the surface image of the test wafer by a certain angle.

[0148] Next, step F100 is performed: acquiring the surface image of the test wafer under the standard image acquisition condition, wherein the surface of the test wafer has a plurality of periodic patterns;

[0149] Next, step F200 is performed: comparing the standard template image with the minimum period images in the surface image of the test wafer one by one;

[0150] Specifically, in the present embodiment, the standard template image is stored in the host computer, and when detection is performed, the standard template image and the standard detection program are directly called, so that the one-by-one measurement of the minimum period images in the surface image of the test wafer can be automatically completed.

[0151] Specifically, in the present embodiment, the test wafer can be conveyed to the support table of the defect detection device by a robot or an automatic production line, the standard detection program is called by the host computer application program, and the one-by-one reading of the minimum period images in the surface image of the test wafer and the one-by-one comparison with the standard template image are completed.

[0152] Further, in the present embodiment, the step of comparing the minimum period images in the standard image with the minimum period images in the surface image of the test wafer comprises:

[0153] performing shape matching using the periodic pattern of the minimum period in the standard image, and selecting the position of the periodic pattern of the minimum period in the surface image of the test wafer that needs to be detected;

[0154] Specifically, in the present embodiment, the method for selecting the position of the minimum period periodic regular pattern to be detected in the surface image of the test wafer can be: mode 1) sampling shape matching at various positions on the surface image of the test wafer to select the position of the minimum period periodic regular pattern to be detected; mode 2) shape matching at various positions on the surface image of the test wafer point by point / row by row / column by column according to a certain scanning rule and scanning moving interval to select the position of the minimum period periodic regular pattern to be detected point by point / row by row / column by column. Preferably, in mode 2), the scanning moving interval can be set as the length or width of the minimum period image according to the direction of scanning movement.

[0155] After the above steps, m minimum period periodic regular patterns to be detected are selected in the surface image of the test wafer, and then compared with the standard template image one by one to perform defect detection.

[0156] Step F300 is performed: judging defects according to the analysis comparison result.

[0157] Specifically, in the present step, the above detection of the minimum period periodic regular pattern by comparing it with the standard template image one by one is based on the difference comparison of the gray level values of the minimum period periodic regular pattern to be detected and the standard template image. When the difference value of the gray level values of the two is greater than the set detection threshold (Threshold), the area where the minimum period periodic regular pattern to be detected is located is determined as a defect area.

[0158] The present concept is described below with specific embodiments.

[0159] In one specific embodiment, it is measured that the gray level (GrayLevel) of the standard template image trained by the clean wafer image is 128, and the standard deviation gray level (SdevGrayLevel) is 1.2.

[0160] The gray level of some minimum period periodic regular pattern areas is 150, and the gray level of the minimum period periodic regular pattern at the remaining positions is 128.

[0161] The set detection threshold (Threshold) is 10, and the standard deviation ratio (SdevRatio) is 3.

[0162] Since 150>128+10+3x1.2=141.6

[0163] Therefore, the area with the image gray level of 150 is determined as a defect area.

[0164] Further, in other specific embodiments, the defect judging includes judging dark defects and bright defects. Specifically as follows:

[0165] 1) The case of judging dark defects is as follows:

[0166] The average map of the standard template image obtained by training the non-polluted wafer image has a gray level of 128, and the standard deviation map has a gray level of 1.2;

[0167] The gray level of the defect detection region is 90;

[0168] The set detection threshold is 15, and the standard deviation ratio is 3;

[0169] Since the gray level of the defect detection region is 90 < 128-15-3x1.2 = 109.4

[0170] The region with the image gray level of 90 is judged as a dark defect region.

[0171] 2) The case of judging bright defects is as follows:

[0172] The average map of the standard template image obtained by training the non-polluted wafer image has a gray level of 128, and the standard deviation map has a gray level of 1.2;

[0173] The gray level of the defect detection region is 150;

[0174] The set detection threshold is 15, and the standard deviation ratio is 3;

[0175] Since the gray level of the defect detection region is 150 > 128+15+3x1.2 = 146.6

[0176] The region with the image gray level of 150 is judged as a bright defect region.

[0177] On the other hand, the technical scheme of the present application also provides a wafer surface defect detection device. Referring to Figure 6 The wafer surface defect detection device includes:

[0178] An image acquisition unit 72 is configured to acquire images of the surface of a wafer 90 (a standard wafer or a test wafer);

[0179] An image storage unit 71 is configured to store the images acquired by the image acquisition unit, and includes a surface image storage area of a standard wafer, a standard image storage area of a standard wafer, and a surface image storage area of a test wafer;

[0180] The comparison unit 73 is configured to compare the image data to be compared or the image data to a preset threshold value;

[0181] The judgment unit 74 is configured to judge the comparison result of the test comparison unit.

[0182] On the basis of the above, in the embodiment, a wafer surface defect detection device is provided, characterized in that comprising:

[0183] An image storage unit is configured to store a standard template image;

[0184] An image acquisition unit is configured to acquire a surface image of a test wafer under standard image acquisition conditions, wherein the surface of the test wafer has a plurality of periodic patterns;

[0185] A test comparison unit is configured to compare the standard template image with each of the minimum period images in the surface image of the test wafer;

[0186] A defect judgment unit is configured to judge defects according to the comparison result of the test comparison unit,

[0187] Wherein,

[0188] When the difference between the standard template image and the minimum period image in the surface image of the test wafer exceeds a judgment threshold value, the region where the compared minimum period image in the surface image of the test wafer is located is judged as a defect region.

[0189] The image acquisition unit is further configured to acquire surface images of standard wafers under different image acquisition conditions, wherein the standard wafers are non-polluted wafers, and the surfaces of the standard wafers have a plurality of periodic patterns similar to the surface of the test wafer;

[0190] The size of the standard template image is related to the minimum period size of the periodic pattern.

[0191] Further, the wafer surface defect detection device provided in the embodiment of the technical scheme of the present application further comprises a standard image conversion unit, comprising:

[0192] A gray value calculation unit is configured to calculate the average gray value of each surface image of the standard wafer acquired under different image acquisition conditions;

[0193] A standard image matching unit is configured to compare the difference between the average gray value of each surface image of the standard wafer and a preset standard gray value, and select the surface image of the standard wafer with an average gray value close to the standard gray value as the standard image of the standard wafer.

[0194] Further, the wafer surface defect detection device provided in the embodiment of the technical scheme of the present application further comprises a standard image acquisition condition setting module, which is adapted to set the environmental condition corresponding to the image of the standard wafer with the average gray value close to the standard value as the standard image acquisition condition.

[0195] Further, the wafer surface defect detection device provided in the embodiment of the technical scheme of the present application further comprises the standard model image processing unit, which comprises:

[0196] An initial template image extraction unit is adapted to extract a periodic regular pattern with the minimum period from the standard image of the standard wafer as an initial template image.

[0197] A pre-template image extraction unit is adapted to extract n periodic regular patterns with the minimum period from the standard image as pre-template images based on the initial template image, the preset matching score and the pre-template image size.

[0198] A matching value screening unit is adapted to eliminate the pre-template images with the highest and lowest matching scores with the initial template image.

[0199] An optimal image synthesis unit is adapted to perform superposition processing on the images of the remaining n-2 pre-template images, obtain an average image and a standard deviation image, and finally combine to obtain an optimal image as the standard model image.

[0200] Further, the wafer surface defect detection device provided in the embodiment of the technical scheme of the present application further comprises:

[0201] A minimum periodic regular pattern extraction unit;

[0202] The minimum periodic regular pattern extraction unit is adapted to: based on the pattern regularity of the surface image of the standard wafer and / or the surface image of the test wafer, use Fourier transform to obtain conjugate coefficients, transform image information from spatial domain to frequency domain, utilize two-dimensional sinusoidal wave to fit the fluctuation of gray value, obtain the periodic regular image size in horizontal and vertical directions, calculate the periodic regular pattern size, and thus extract the periodic regular pattern with the minimum period from the surface of the wafer.

[0203] Further, the wafer surface defect detection device provided in the embodiment of the technical scheme of the present application further comprises a surface image initial positioning unit of the test wafer, which is adapted to search the position area of the wafer positioning groove in the surface image of the test wafer and judge whether the position of the wafer positioning groove is accurate.

[0204] If not, rotate the surface image of the test wafer by a certain angle.

[0205] Further, the wafer surface defect detection device provided in the embodiment of the technical solution of the present application, the test comparison module comprises:

[0206] The shape matching unit is adapted to perform shape matching by using the periodic regular pattern with the smallest period in the standard image, and select the position of the periodic regular pattern with the smallest period in the surface image of the test wafer which needs to be detected;

[0207] The difference unit is adapted to perform difference between the periodic regular pattern with the smallest period in the surface image of the test wafer which needs to be detected and the average image.

[0208] Further, the wafer surface defect detection device provided in the embodiment of the technical solution of the present application comprises:

[0209] The processor,

[0210] The memory stores computer programs or instructions, which, when executed by the processor, implement the steps of the method according to any one of the above embodiments to perform defect detection on the object.

[0211] Further, the embodiment of the technical solution of the present application further provides a computer program product, which is stored in a computer-readable storage medium and comprises computer programs or instructions, which, when executed by the processor, implement the steps of the method.

[0212] The above embodiments are only for illustrating the technical concept and characteristics of the present application, and the purpose is to enable those skilled in the art to understand the content of the present application and implement it, but it is not used to limit the protection scope of the present application.

[0213] Any person skilled in the art can make possible changes and modifications to the technical solution of the present application without departing from the spirit and scope of the present application by using the above disclosed methods and technical contents, therefore, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present application without departing from the content of the technical solution of the present application, all belong to the protection scope of the technical solution of the present application.

Claims

1. A method for detecting defects on a wafer surface, characterized in that, include: Acquire a surface image of a test wafer under standard image acquisition conditions, wherein the test wafer has undergone photoresist coating and development, and its surface has a periodically and regularly arranged chip pattern; Obtain the standard template image corresponding to the test wafer, the set detection threshold and the standard deviation multiple, wherein the standard template image includes a minimum periodic regular pattern and the standard template image includes gray values ​​and standard deviation gray values; The minimum period images in the surface images of the test wafer are matched one by one using a standard template image to select the minimum period image to be detected. Based on the set detection threshold, the standard deviation multiple, the gray value, and the standard deviation gray value, the bright defect threshold and the dark defect threshold are calculated respectively, and the bright defect threshold is greater than the dark defect threshold. When the gray value of the minimum period image to be detected is greater than the bright defect threshold, the region where the gray value is located is determined to be a bright defect region. When the gray value of the minimum period image to be detected is less than the dark defect threshold, the region where the gray value is located is determined to be a dark defect region. The steps for obtaining the standard template image include: Based on the graphic patterns of the wafer image, Fourier transform is used to find the conjugate coefficients, transforming the image information from the spatial domain to the frequency domain. The fluctuations of gray values ​​are fitted using a two-dimensional sine wave to determine the size of the periodic pattern image in the horizontal and vertical directions. The size of the periodic pattern is calculated to obtain the smallest periodic pattern image, and the smallest periodic pattern image is determined as the initial template image. Based on the initial template image, a pre-set matching score, and the size of the preceding template image, a periodic pattern with n minimum cycles is extracted from the standard image as the preceding template image. Remove the preceding template images that have the highest and lowest matching scores with the initial template image; The remaining n-2 preceding template images are overlaid to obtain an average image and a standard variance image, and finally merged to obtain the best image as the standard template image. The step of performing shape matching on the minimum period images in the surface images of the test wafer one by one using a standard template image includes: Shape matching is performed by sampling various locations on the surface image of the test wafer to select the position of the periodic pattern with the smallest detected period; or, On the surface image of the test wafer, according to a preset scanning pattern and scanning movement interval, shape matching is performed point by point / row by column on each part of the surface image of the test wafer, and the position of the periodic pattern with the smallest period detected is selected point by point / row by column.

2. The method for detecting wafer surface defects according to claim 1, characterized in that, The steps for obtaining the standard template image include: Under standard acquisition conditions, a surface image of a standard wafer is obtained to obtain a standard image of the standard wafer; The standard template image is obtained by processing the standard image of the standard wafer; The standard wafer is a pollution-free wafer, and its surface has several periodic patterns similar to those on the surface of the test wafer. The size of the standard template image is related to the minimum period size of the periodic pattern.

3. The method for detecting wafer surface defects according to claim 2, characterized in that, The steps for obtaining the standard image acquisition conditions include: Images of a standard wafer were acquired under different image acquisition conditions; Calculate the average grayscale value for each image; The environmental conditions corresponding to the image with the smallest average gray value that differs from the standard value are taken as the standard image acquisition conditions.

4. The method for detecting wafer surface defects according to claim 1, characterized in that, After the step of acquiring the surface image of the test wafer under standard image acquisition conditions, and before the step of comparing the standard template image with the smallest period image in the surface image of the test wafer one by one, the method further includes: performing initial positioning of the surface image of the test wafer.

5. The method for detecting wafer surface defects according to claim 4, characterized in that, The step of initially locating the surface image of the test wafer includes: Search for the location area of ​​the wafer positioning groove in the surface image of the test wafer; Determine if the position of the wafer positioning groove is accurate; If inaccurate, rotate the surface image of the test wafer by a preset angle.

6. The method for detecting wafer surface defects according to claim 3, characterized in that, The different image acquisition conditions include different settings for factors such as brightness and darkness, LED brightness, camera exposure parameters, and / or camera gain.

7. A device for detecting defects on the surface of a wafer, characterized in that, include: The standard model image processing unit is used to: based on the graphic patterns of the wafer image, use Fourier transform to find the conjugate coefficients, transform the image information from the spatial domain to the frequency domain, use a two-dimensional sine wave to fit the fluctuations of gray values, find the dimensions of the periodic patterns in the horizontal and vertical directions, calculate the dimensions of the periodic patterns to obtain the minimum periodic patterns, and determine the minimum periodic patterns as the initial template image; based on the initial template image, preset matching scores and the dimensions of the previous template image, extract n minimum periodic patterns from the standard image as the previous template images; Remove the preceding template images with the highest and lowest matching scores to the initial template image; overlay the remaining n-2 preceding template images to obtain the average image and standard variance image, and finally merge them to obtain the best image as the standard template image. An image storage unit is adapted to store a standard template image, a set detection threshold, and a standard variance multiple, wherein the standard template image includes grayscale values ​​and standard variance grayscale values; an image acquisition unit is adapted to acquire a surface image of a test wafer under standard image acquisition conditions, wherein the test wafer has undergone photoresist coating and development, and its surface has a periodically arranged chip pattern. The test comparison unit is adapted to perform shape matching on the minimum period images in the surface image of the test wafer one by one using a standard template image to select the minimum period image to be detected. The shape matching on the minimum period images in the surface image of the test wafer using the standard template image includes: sampling and performing shape matching at various points on the surface image of the test wafer to select the position of the periodic pattern of the minimum period to be detected; or, performing shape matching on various points on the surface image of the test wafer according to a preset scanning pattern and scanning movement interval, point-by-point / row-by-column, to select the position of the periodic pattern of the minimum period to be detected; and calculating a bright defect threshold and a dark defect threshold based on the set detection threshold, the standard deviation multiple, the gray value, and the standard deviation gray value, wherein the bright defect threshold is greater than the dark defect threshold. The defect judgment unit is adapted to judge defects based on the bright defect threshold and dark defect threshold of the test comparison unit. in, When the gray value of the minimum period image to be detected is greater than the bright defect threshold, the region where the gray value is located is determined to be a bright defect region. When the gray value of the minimum periodic image to be detected is less than the dark defect threshold, the region where the gray value is located is determined to be a dark defect region.

8. The apparatus for detecting wafer surface defects according to claim 7, characterized in that, The image acquisition unit is also adapted to acquire surface images of a standard wafer under different image acquisition conditions, wherein the standard wafer is a contamination-free wafer and its surface has several periodic patterns similar to those on the surface of the test wafer. The size of the standard template image is related to the minimum period size of the periodic pattern.

9. The apparatus for detecting wafer surface defects according to claim 7, characterized in that, It also includes a standard image conversion unit, including: The grayscale calculation unit is suitable for calculating the average grayscale value of each surface image of a standard wafer acquired under different image acquisition conditions. The standard image matching unit is adapted to compare the difference between the average gray value of each standard wafer surface image and the preset standard gray value, and select the surface image of the standard wafer with the smallest average gray value difference from the standard gray value as the standard image of the standard wafer.

10. The apparatus for detecting wafer surface defects according to claim 9, characterized in that, Also includes: The standard image acquisition condition setting module is suitable for setting the environmental conditions corresponding to the standard image to which the average gray value with the smallest difference from the standard value belongs as the standard image acquisition condition.

11. The apparatus for detecting wafer surface defects according to claim 8, characterized in that, It also includes a surface image initial positioning unit for the test wafer, which is suitable for searching the location area of ​​the wafer positioning groove in the surface image of the test wafer and determining whether the position of the wafer positioning groove is accurate; If inaccurate, rotate the surface image of the test wafer by a preset angle.

12. An apparatus for detecting defects on the surface of a wafer, characterized in that, include: processor, A memory that stores a computer program or instructions that, when executed by a processor, implement the steps of the method according to any one of claims 1 to 6, for defect detection of an object.

13. A computer program product, characterized in that, The computer program product is stored on a computer-readable storage medium and includes a computer program or instructions that, when executed by a processor, implement the steps of the method according to any one of claims 1 to 6.

Images