A method for assisting in improving product particle detection in semiconductor processing

Abstract

The application relates to the technical field of semiconductor manufacturing detection, and discloses a method for assisting in improving product particle detection in semiconductor processing. The method acquires a particle detection result and a corresponding detection image of a particle detection device on a semiconductor product to be detected, determines a product to be judged with a particle defect and a color fog mode, selects a detection sample therefrom, and adopts a non-contact optical surface topography detection device to perform roughness detection, so as to obtain a roughness detection value; a preset roughness specification range corresponding to the product to be judged is acquired, the roughness detection value is compared with the specification range, when the roughness detection value is within the specification range, it is determined that the product meets normal flow conditions, and when the roughness detection value exceeds the specification range, the product is transferred to a polishing process for rework processing. The application can reduce unnecessary rework caused by the color fog mode through the roughness detection result to assist in particle detection judgment.

Description

Technical Field

[0001] This invention relates to the field of semiconductor manufacturing inspection technology, and in particular to a method for improving particle detection in semiconductor processing. Background Technology

[0002] During the processing of semiconductor products, it is usually necessary to inspect the particle condition on the product surface to determine whether there are any abnormalities such as excessive particle count. The particle inspection results directly affect whether the product can continue to the subsequent processing steps. Therefore, the accurate determination of particle inspection results is of great significance for improving the production efficiency of semiconductor products and ensuring product quality.

[0003] In existing semiconductor processing, particle inspection equipment can be used to inspect the surface of the processed product. However, in actual inspection, due to factors such as the surface condition of the product and the processing technology, colored fog may appear on the product surface. Colored fog is a normal surface phenomenon in some semiconductor processing, but it may appear simultaneously with abnormal particle detection results in the detection image generated by the particle inspection equipment.

[0004] When a particle detection device detects defective particles and a colored fog pattern appears in the corresponding detection image, it is difficult to determine whether the abnormal detection result is caused by an actual particle over-limit or by the colored fog pattern affecting the detection process based solely on the detection result output by the particle detection device. Therefore, in the above situation, it is easy to encounter the problem of inaccurate particle detection results.

[0005] In the existing production process, products that are detected as defective due to colored fog mode by particle detection equipment are usually returned to the polishing process for rework. Although rework can reduce the risk of defective products continuing to flow, reworking products that actually meet the product specifications but have abnormal detection results only due to colored fog mode will increase unnecessary processing steps. This not only reduces production efficiency, but also may increase the risk of new processing defects after the products are polished again.

[0006] Therefore, relying solely on the detection results of the particle detection equipment itself is insufficient to effectively address the issue of further determining defective particle detection results under colored fog mode. It is necessary to provide a detection method that can introduce physical parameters of the product surface for further auxiliary judgment based on the detection results of the particle detection equipment, in order to distinguish between products that require rework and products that can flow normally, thereby improving the reliability of particle detection results under colored fog mode and reducing unnecessary rework. Summary of the Invention

[0007] To address the technical problems existing in the prior art, embodiments of the present invention provide a method for assisting in improving particle detection in semiconductor processing. The technical solution is as follows: A method for improving particle detection in semiconductor processing includes the following steps: S1: Obtain the particle detection results and the corresponding detection image obtained by the particle detection equipment from the particle detection of the semiconductor product to be tested; S2: Based on the particle detection results and the detection image, determine the product to be judged that the particle detection results indicate the presence of particle defects and the detection image shows a color fog mode; S3: Select a test sample from the products to be judged, and use a roughness testing device to perform roughness testing on the test surface of the test sample to obtain the roughness test value of the test surface. S4: Obtain a preset roughness specification range corresponding to the product to be judged, and compare the roughness detection value with the preset roughness specification range; S5: Based on the comparison result between the roughness detection value and the preset roughness specification range, the particle detection result is auxiliaryly judged, wherein when the roughness detection value is within the preset roughness specification range, it is determined that the product to be judged meets the normal flow conditions. When the roughness detection value exceeds the preset roughness specification range, it is determined that the product to be judged does not meet the normal flow conditions, and the product to be judged is reworked.

[0008] As a further aspect of the present invention, the step of determining a product to be judged that indicates the presence of defective particles in the particle detection result and whose detection image presents a colored fog pattern includes: acquiring particle abnormality detection information output by the particle detection device, determining a detection image corresponding to the particle abnormality detection information, and identifying whether colored fog features exist in the detection image. When both the particle anomaly detection information and the colored fog feature are present, the corresponding semiconductor product is identified as a product to be judged that requires roughness auxiliary detection.

[0009] As a further aspect of the present invention, the roughness detection device is a non-contact optical surface morphology detection device; The roughness detection of the test surface of the test sample using a roughness detection device includes acquiring surface morphology information of the test surface and obtaining at least one roughness parameter to characterize the roughness of the test surface based on the surface morphology information.

[0010] As a further aspect of the present invention, when performing roughness testing on the test sample, at least one roughness testing area is determined on the surface to be tested, and roughness testing values ​​corresponding to each roughness testing area are obtained respectively. When multiple roughness detection areas are provided, a roughness characterization value for the auxiliary determination is obtained based on the roughness detection values ​​of the multiple roughness detection areas.

[0011] As a further aspect of the present invention, the roughness detection area is determined based on the detection image obtained by the particle detection device, and includes at least a detection area corresponding to the colored fog area in the detection image, so that the roughness detection result corresponds to the surface state of the colored fog area.

[0012] As a further aspect of the present invention, the preset roughness specification range is determined according to the customer specifications corresponding to the product to be judged. Before performing the auxiliary judgment, a corresponding preset roughness specification range is obtained based on the product specification information of the product to be judged, and the preset roughness specification range is used as the judgment criterion for the roughness detection value.

[0013] As a further aspect of the present invention, selecting test samples from the products to be judged includes extracting at least one product from the products that are determined by the particle detection equipment to have colored fog mode particle defects as the test samples. The roughness of the test sample is tested, and the roughness test results are used to determine whether the corresponding product to be judged meets the normal flow conditions.

[0014] As a further embodiment of the present invention, when the roughness detection value is within the preset roughness specification range, the pending judgment state formed by the color fog mode and the particle defect detection result is released, so that the corresponding product can enter the subsequent processing flow. When the roughness detection value exceeds the preset roughness specification range, the corresponding product will be transferred to the polishing process for rework.

[0015] The beneficial effects of the technical solutions provided by the embodiments of the present invention include at least the following: This invention further introduces product surface roughness detection based on the detection results of particle detection equipment. By forming a two-level auxiliary judgment mechanism through particle detection results, color fog mode, and surface roughness detection results, it solves the problem that particle detection equipment has difficulty in directly judging whether particle defects need to be reworked when a color fog mode is present.

[0016] This invention does not directly return the product to the polishing process after the particle detection equipment detects defective particles in the colored fog mode. Instead, it further obtains the roughness detection value of the product to be judged and compares the roughness detection value with the preset roughness specification range of the corresponding product, so that the flow judgment of the product has a further objective detection basis.

[0017] This invention uses the roughness specification range determined by the customer's specifications for the corresponding product as an auxiliary judgment standard, so that the roughness auxiliary test results can correspond to the acceptance requirements of the specific product. When the roughness test value meets the specified range, the product can proceed normally into the subsequent processing flow, thereby reducing unnecessary rework caused by the colored fog mode.

[0018] For products whose roughness test values ​​exceed the preset specification range, the present invention still transfers them to rework processing. This reduces unnecessary rework while retaining quality control measures for actual abnormal products, which is beneficial to balancing production efficiency and product quality.

[0019] By reducing the number of products that meet the surface roughness specifications from re-entering the polishing process, the number of repeated processing steps can be reduced, the handling process for defective products can be shortened, the processing efficiency of semiconductor products can be improved, and the risk of generating new surface defects or other processing defects during secondary processing can be reduced. Attached Figure Description

[0020] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0021] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0022] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0023] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0024] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0025] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0026] Please see Figure 1 This invention provides a technical solution: a method for improving particle detection in semiconductor processing, characterized by comprising the following steps: S1: Acquire the particle detection results and the corresponding detection images obtained by the particle detection equipment from particle detection of the semiconductor product to be tested; S2: Based on the particle detection results and the detection images, identify the products to be judged that indicate the presence of particle defects in the particle detection results and show a color fog mode in the detection images; S3: Select a test sample from the products to be judged, and use a roughness testing device to test the roughness of the test sample's surface to obtain the roughness test value of the test surface. S4: Obtain the preset roughness specification range corresponding to the product to be judged, and compare the roughness detection value with the preset roughness specification range; S5: Based on the comparison between the roughness detection value and the preset roughness specification range, the particle detection result is used for auxiliary judgment. When the roughness detection value is within the preset roughness specification range, it is determined that the product to be judged meets the normal flow conditions. When the roughness test value exceeds the preset roughness specification range, it is determined that the product to be judged does not meet the normal flow conditions, and the product to be judged is reworked.

[0027] Further, determining the product to be judged, where the particle detection result indicates the presence of particle defects and the detection image presents a colored fog pattern, includes: acquiring particle abnormality detection information output by the particle detection equipment, determining the detection image corresponding to the particle abnormality detection information, and identifying whether colored fog features exist in the detection image; When both particle anomaly detection information and colored fog characteristics are present, the corresponding semiconductor product is identified as a product to be judged that requires roughness auxiliary detection. Starting with a single-piece inspection record generated by the particle detection equipment, the product number, scanning time, particle anomaly identifier, and anomaly coordinates are read. The corresponding inspection image is retrieved based on the same product number and scanning time. The gray levels of the R, G, and B channels are read pixel by pixel, and the pixel color difference is obtained by subtracting the minimum channel gray level from the maximum channel gray level. The lower limit of the color difference of 15 gray levels of 20 manually confirmed colored fog samples of the same product is taken as the color difference threshold. Adjacent pixels with a color difference of 15 gray levels are connected and the area of ​​the continuous region is counted. The minimum area of ​​the continuous region of the aforementioned 20 samples, 0.8 square millimeters, is taken as the area threshold. For example, if the inspection record of a certain product shows that the particle anomaly identifier is FAIL, the corresponding image forms a continuous region of 1.6 square millimeters with a pixel color difference of 18 to 42 gray levels. This product is registered as a roughness verification object.

[0028] Furthermore, the surface roughness testing equipment is a non-contact optical surface morphology testing equipment; Roughness testing is performed on the surface of the test sample using a roughness testing device, including acquiring surface morphology information of the surface to be tested, and obtaining at least one roughness parameter to characterize the roughness of the surface to be tested based on the surface morphology information. Based on a non-contact optical surface morphology inspection device, the colored haze wafer to be judged detected by SP-1 is fixed on the inspection stage. The measurement configuration corresponding to the product number is called, the inspection area is set to 1 mm × 1 mm, and the sampling interval is set to 0.5 micrometers. After optical focusing is completed, the spatial height information of the surface to be measured is collected point by point. Sampling points marked by the device as out of focus or without return signal are removed. The remaining sampling points are arranged in their original spatial positions to form a surface morphology record. For example, after obtaining continuous effective sampling points in a certain wafer inspection area, the device reads the roughness characterization result corresponding to that area and records the roughness parameters corresponding to the product number and inspection area.

[0029] Furthermore, when performing roughness testing on the test sample, at least one roughness testing area is determined on the surface to be tested, and the roughness testing value corresponding to each roughness testing area is obtained respectively. When multiple roughness detection areas are set, roughness characterization values ​​used for auxiliary judgment are obtained based on the roughness detection values ​​of multiple roughness detection areas. Based on the roughness detection area settings of the surface to be tested, the position corresponding to the colored fog area detected by SP-1 is set as the first detection area, and the second and third detection areas are set at the adjacent positions. The non-contact optical surface morphology detection equipment is called in sequence to complete the scanning of each area and read the roughness parameters output by the equipment. For example, the roughness parameters of the three areas are 0.18nm, 0.22nm and 0.20nm respectively. The three values ​​are added together to get 0.60nm and divided by the number of detection areas 3 to get 0.20nm. This value is registered in the detection record of the corresponding product number to obtain the roughness characterization value used for auxiliary judgment.

[0030] Furthermore, the roughness detection area is determined based on the detection image obtained by the particle detection equipment, and includes at least a detection area corresponding to the colored fog area in the detection image, so that the roughness detection result corresponds to the surface condition of the colored fog area; Based on particle detection images, a correspondence between wafer surface positions is established. The product number, image center coordinates, boundary coordinates of the continuous colored fog region, and equipment scanning orientation are read. The image center is set as the wafer center reference point. The colored fog boundary position is mapped to the detection stage according to the same scanning orientation. For example, if the colored fog is located 12.0 mm to the right of the center and 8.0 mm above it, and the area is 1.6 mm², then a 1 mm × 1 mm area is selected at the same offset position on the sample surface and the roughness scan is completed. The correspondence between the roughness result of this area and the position of the colored fog is recorded.

[0031] Furthermore, the preset roughness specification range is determined based on the customer specifications corresponding to the product to be judged; Before making auxiliary judgments, the corresponding preset roughness specification range is obtained based on the product specification information of the product to be judged, and the preset roughness specification range is used as the judgment benchmark for the roughness detection value. Based on the customer specification record corresponding to the product to be judged, the product number, customer code, specification version number, roughness item and its upper and lower limits are read. The product number is checked against the part number in the specification file item by item, and then the roughness limit under the same version is called. For example, if product A corresponds to customer specification V3.2 with a registered lower limit of Ra of 0.10nm and an upper limit of 0.30nm, then 0.10nm to 0.30nm is registered as the specification range of the product. When 0.20nm is detected, it is directly checked whether 0.20nm is not less than 0.10nm and not greater than 0.30nm at the same time, and this specification range is registered as the judgment criterion for the roughness test result of the product.

[0032] Furthermore, test samples are selected from the products to be judged, including at least one product selected as a test sample from products that are determined by the particle detection equipment to have colored fog pattern particle defects. The roughness of the test sample is tested, and the results of the roughness test are used to determine whether the corresponding product to be judged meets the normal flow conditions. Based on the list of products to be judged that are marked as abnormal colored fog particles by SP-1, the product number, abnormality mark and detection time of each product in the same batch are read. At least one piece is randomly selected as a test sample in numerical order. For example, if 4 out of 25 products in batch A are marked as abnormal, then product A08 is selected for ZYGO roughness detection. The Ra results of 0.18nm, 0.21nm and 0.20nm in the corresponding detection area are read and checked against the customer specifications of 0.10nm to 0.30nm item by item. If all three results are between the lower limit and the upper limit, they are registered as within the specification. If any result is lower than 0.10nm or higher than 0.30nm, it is registered as outside the specification. Based on this, it is determined whether the corresponding product to be judged meets the normal flow conditions.

[0033] Furthermore, when the roughness detection value is within the preset roughness specification range, the pending judgment state formed by the color fog mode and the particle defect detection results is released, allowing the corresponding product to enter the subsequent processing flow. When the roughness test value exceeds the preset roughness specification range, the corresponding product will be transferred to the polishing process for rework. Based on the roughness inspection records of the products to be judged, the product number, inspection area, Ra result, and corresponding upper and lower limits of customer specifications are read. The Ra result is checked against the lower limit of 0.10nm and the upper limit of 0.30nm item by item. For example, if the Ra of product A08 is 0.20nm, it is registered as within the specification and the judgment mark is changed to normal flow. If the Ra of product A12 is 0.34nm, it is registered as outside the specification and a rework mark is generated. At the same time, the product number, abnormal value, and rework process code are recorded, and the corresponding product is transferred to the polishing process for rework processing.

[0034] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for improving particle detection in semiconductor processing, characterized in that, Includes the following steps: S1: Obtain the particle detection results and the corresponding detection image obtained by the particle detection equipment from the particle detection of the semiconductor product to be tested; S2: Based on the particle detection results and the detection image, determine the product to be judged that the particle detection results indicate the presence of particle defects and the detection image shows a color fog mode; S3: Select a test sample from the products to be judged, and use a roughness testing device to perform roughness testing on the test surface of the test sample to obtain the roughness test value of the test surface. S4: Obtain a preset roughness specification range corresponding to the product to be judged, and compare the roughness detection value with the preset roughness specification range; S5: Based on the comparison result between the roughness detection value and the preset roughness specification range, the particle detection result is auxiliaryly judged, wherein when the roughness detection value is within the preset roughness specification range, it is determined that the product to be judged meets the normal flow conditions. When the roughness detection value exceeds the preset roughness specification range, it is determined that the product to be judged does not meet the normal flow conditions, and the product to be judged is reworked.

2. The method for assisting in improving particle detection in semiconductor processing according to claim 1, characterized in that: The step of determining a product to be judged that indicates the presence of defective particles in the particle detection result and whose detection image presents a colored fog pattern includes: acquiring particle abnormality detection information output by the particle detection device, determining a detection image corresponding to the particle abnormality detection information, and identifying whether colored fog features exist in the detection image. When both the particle anomaly detection information and the colored fog feature are present, the corresponding semiconductor product is identified as a product to be judged that requires roughness auxiliary detection.

3. The method for assisting in improving particle detection in semiconductor processing according to claim 1, characterized in that: The roughness detection equipment is a non-contact optical surface morphology detection equipment; The roughness detection of the test surface of the test sample using a roughness detection device includes acquiring surface morphology information of the test surface and obtaining at least one roughness parameter to characterize the roughness of the test surface based on the surface morphology information.

4. The method for assisting in improving particle detection in semiconductor processing according to claim 3, characterized in that: When performing roughness testing on the test sample, at least one roughness testing area is determined on the surface to be tested, and the roughness testing value corresponding to each roughness testing area is obtained respectively. When multiple roughness detection areas are provided, a roughness characterization value for the auxiliary determination is obtained based on the roughness detection values ​​of the multiple roughness detection areas.

5. The method for assisting in improving particle detection in semiconductor processing according to claim 4, characterized in that: The roughness detection area is determined based on the detection image obtained by the particle detection device, and includes at least a detection area corresponding to the colored fog area in the detection image, so that the roughness detection result corresponds to the surface state of the colored fog area.

6. The method for assisting in improving particle detection in semiconductor processing according to claim 1, characterized in that: The preset roughness specification range is determined according to the customer specifications corresponding to the product to be judged. Before performing the auxiliary judgment, a corresponding preset roughness specification range is obtained based on the product specification information of the product to be judged, and the preset roughness specification range is used as the judgment criterion for the roughness detection value.

7. The method for assisting in improving particle detection in semiconductor processing according to claim 1, characterized in that: Selecting test samples from the products to be judged includes extracting at least one product from the products that are determined by the particle detection equipment to have colored fog mode particle defects as the test samples; The roughness of the test sample is tested, and the roughness test results are used to determine whether the corresponding product to be judged meets the normal flow conditions.

8. The method for assisting in improving particle detection in semiconductor processing according to claim 1, characterized in that: When the roughness detection value is within the preset roughness specification range, the pending judgment state formed by the color fog mode and the particle defect detection result is released, allowing the corresponding product to enter the subsequent processing flow. When the roughness detection value exceeds the preset roughness specification range, the corresponding product will be transferred to the polishing process for rework.

Images