Wafer surface defect level determination method

Abstract

The application discloses a wafer surface defect grade determination method, which comprises the following steps: obtaining a developing imaging picture containing a wafer to be measured and determining an imaging diameter of an outer edge thereof; determining the size of a pixel value corresponding to the imaging diameter of the outer edge and a pixel value corresponding to a wafer specification diameter, so as to determine whether the actual outer edge of the wafer is within the imaging picture outer edge; if the actual outer edge of the wafer is within the imaging picture outer edge, cutting the area outside the outer edge of the wafer in the imaging picture; dividing a plurality of defect determination areas in the imaging picture; determining the number of each defect shape in each defect determination area, so as to determine the defect grade of the wafer. The application can accurately calculate the real area of the wafer to be measured, quickly lock the area range of the wafer to be measured, determine the positions of different defects based on the defect shapes, comprehensively determine the defect grade of the wafer, and provide accurate defect analysis for production.

Description

Technical Field

[0001] This invention belongs to the field of semiconductor silicon wafer defect detection technology, and in particular relates to a method for determining the level of surface defects on wafers. Background Technology

[0002] The principle underlying the PL defect tester is photoluminescence. It utilizes a laser of a specific wavelength as the excitation source, providing photons of a certain energy. Electrons in the sample, currently in the ground state, absorb these photons and enter the excited state. These excited electrons are metastable and will return to the ground state within a short time, emitting fluorescence with an infrared peak around 1150nm (using a SI cell as an example). A high-sensitivity, high-resolution camera is used to capture the light, and the images are then analyzed by software.

[0003] The intensity of emission is directly proportional to the concentration of non-equilibrium minority carriers at that location. Defects act as strong recombination centers for minority carriers, thus reducing the minority carrier concentration in that region and weakening the fluorescence effect. This manifests as dark dots, lines, or specific areas on the image, while areas with less recombination appear as brighter areas. Therefore, by observing photoluminescence imaging (PL), one can determine whether the sample contains defects, impurities, and other factors that ultimately affect battery efficiency.

[0004] Chinese patent CN201680060750.5 proposes a method for determining defect areas in silicon wafers cut from single-crystal silicon manufactured by the CZ method. This method primarily addresses the principle of defect area determination in silicon single crystals grown while reducing the pulling speed, mainly employing non-destructive inspection using a particle counter. However, this method is not only complex but also unsuitable for various wafer types, especially for wafers where defects are not affected by pulling speed, failing to accurately determine defect levels and classifications.

[0005] The current standard for defect assessment involves taking images with a PL (Portable Layout) device, followed by manual assessment of the defect level based on experience, and then recording the results into a terminal device for storage. However, manual assessment is subject to many interfering factors, incurs high costs, and is inefficient. Summary of the Invention

[0006] This invention provides a method for determining the surface defect level of a wafer, which is particularly suitable for intelligent determination of wafer defects.

[0007] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0008] A method for determining the surface defect level of a wafer, comprising the following steps:

[0009] The pixel value corresponding to the actual diameter of the wafer under test is compared with the pixel value corresponding to the specified diameter of the wafer to determine whether the imaging outline of the wafer is within the outline of the developed imaging image of the wafer.

[0010] If the imaging contours of the wafers are all within the contours of the imaging image, then the area outside the imaging contours corresponding to the wafers in the imaging image is cropped.

[0011] Then, divide the actual outline of the wafer into several regions;

[0012] The defect level of the wafer is determined based on the projection shape of each defect within the imaging profile corresponding to the wafer in the imaging image and the number of defects in different regions.

[0013] Furthermore, the step of determining whether the imaging outline of the wafer is within the outline of the developed image of the wafer, by comparing the pixel value corresponding to the actual diameter of the wafer under test with the pixel value corresponding to the wafer's specified diameter, includes:

[0014] The pixel reference coefficient is obtained based on the ratio of the outer edge imaging diameter of the image to its pixel value;

[0015] Based on the pixel reference coefficient, the pixel values ​​corresponding to the actual diameter and the specified diameter are obtained respectively;

[0016] Then compare the pixel value corresponding to the actual diameter with the pixel value corresponding to the specified diameter.

[0017] Furthermore, when the pixel value corresponding to the actual diameter is less than the pixel value corresponding to the specified diameter;

[0018] Therefore, the imaging contour of the wafer is outside the imaging image contour;

[0019] Adjust the camera focal length and re-develop the image until the pixel value corresponding to the actual diameter is greater than the pixel value corresponding to the specified diameter.

[0020] Furthermore, the step of further dividing the actual outline of the wafer into several regions includes:

[0021] The wafer is divided into concentric inner and outer circles within its actual outline.

[0022] The area occupied by the inner circle is the inner ring area;

[0023] The annular region between the inner circle and the outer circle is the outer ring region; and

[0024] The annular region between the outer circle and the actual contour of the wafer is the outer edge region.

[0025] Furthermore, the inner circle diameter is half the outer circle diameter, and the outer circle diameter is the specified diameter.

[0026] Furthermore, the specified diameter is smaller than the actual diameter of the wafer.

[0027] Furthermore, the step of determining the defect level of the wafer also includes determining the development color within the imaging outline corresponding to the wafer in the imaging image based on each type of defect;

[0028] The defects are rendered in the image in shades of gray and black.

[0029] As the number of defects of the specified shape decreases in the region, the defect level of the wafer increases sequentially.

[0030] Furthermore, the shape of the defect within the imaging profile corresponding to the wafer includes:

[0031] One defect of the elongated, arc-shaped structure;

[0032] The second defect is a ring-shaped structure with uneven width; and

[0033] The third defect of the elliptical structure;

[0034] The first defect, the second defect, and the third defect do not overlap.

[0035] Furthermore, the defect is distributed in the outer edge region;

[0036] The second defect is distributed in the inner ring area and the outer ring area;

[0037] The three defects are distributed in the inner ring area.

[0038] Furthermore, the criteria for determining the defect level of the wafer include, in order:

[0039] The number of defects in the outer edge region;

[0040] The quantity and color of defect 2 in the inner ring area and / or the outer ring area; and

[0041] The number of defect three in the inner ring area;

[0042] Wherein, the number of defects in the outer edge region is not less than two;

[0043] The color of the second defect in the inner ring area and / or the outer ring area is selected as black and gray, respectively.

[0044] The wafer surface defect level determination method designed by this invention can accurately calculate the true area of ​​the tested sample and quickly locate the region of the tested sample; then, based on the defect shape and the color of the imaging position, the location of different defects is determined, and the defect level of the wafer is comprehensively determined so as to provide accurate defect analysis for production. Attached Figure Description

[0045] Figure 1 This is a schematic diagram of the structure of a wafer and an imaging image according to an embodiment of the present invention;

[0046] Figure 2 This is a schematic diagram of the structure of a defect in the imaging profile according to an embodiment of the present invention.

[0047] In the picture:

[0048] 10. Wafer; 11. Outer edge region; 12. Outer ring region

[0049] 13. Inner Ring Area; 20. Imaging Image; 30. Imaging Outline

[0050] 40, Defect 1; 50, Defect 2; 60, Defect 3 Detailed Implementation

[0051] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.

[0052] This embodiment proposes a method for determining the surface defect level of a wafer, such as... Figure 1 As shown, the steps include:

[0053] S1. Determine the size of the pixel value X1 corresponding to the actual diameter D1 of the wafer 10 being tested and the pixel value X2 corresponding to the nominal diameter D2 of the wafer 10, so as to determine whether the imaging outline 30 of the wafer 10 is within the outline of the developed imaging image 20 of the wafer.

[0054] The imaging process involves using a camera to capture an image of the wafer 10 to be tested, which is placed on a table (see attached diagram). To ensure that the wafer 10 is completely captured by the camera lens, the outline of the image 20 containing the wafer 10 is generally larger than the image outline 30 of the wafer 10's projection in the image 20. To ensure the accuracy of the inspection machine in identifying the location of the projected shape of defects within the wafer 10 in the image 20, the area of ​​the wafer 10 outside the image outline 30 in the image 20 needs to be cropped.

[0055] Before determining the size of the imaging contour 30 of the wafer 10 in the imaging image 20, it is necessary to first determine the developed imaging image 20 containing the wafer 10 under test, and determine the outer edge imaging diameter D0 of its contour. After selecting a camera, adjusting its focal length will determine the outer edge imaging diameter D0 of the contour in the imaging image 20 and its corresponding pixel value X0.

[0056] Based on data obtained from multiple experiments by the applicant, it was found that there is a certain relationship between the outer edge imaging diameter D0 of the camera and its pixel value X0, as well as between the actual diameter D1 of the wafer 10 and its corresponding pixel value X1, and between the nominal diameter D2 and its corresponding pixel value X2. That is, the ratio of the outer edge imaging diameter D0 to its pixel value X0 is consistent with the ratio of the actual diameter D1 of the wafer 10 to its corresponding pixel value X1, and the ratio of the nominal diameter D2 to its corresponding pixel value X2. Accordingly, the ratio of diameter to pixel value is used as the pixel reference coefficient ω, with units of mm / pixel.

[0057] That is, based on the ratio of the outer edge imaging diameter D0 to its pixel value X0 in the imaging image 20, the pixel reference coefficient ω is obtained; then based on the pixel reference coefficient ω and the actual diameter D1 and the specification diameter D2 of the wafer 10, the pixel value X1 corresponding to the actual diameter D1 and the pixel value X2 corresponding to the specification diameter D2 are obtained respectively; then the pixel value X1 corresponding to the actual diameter D1 and the pixel value X2 corresponding to the specification diameter D2 are compared to accurately determine that the imaging shape of the defect falls completely within the imaging outline 30 of the wafer 10 and does not include the background pattern on the table.

[0058] Assuming the camera has 1024 pixels X0 and an outer imaging diameter D0 of 330 mm, then ω is approximately 0.322 mm / pixel.

[0059] If the diameter D2 of wafer 10 is 295mm, then its corresponding pixel value X2 is 295 / 0.3222≈915 pixels.

[0060] In actual production, the outer edge of the pulled single crystal, i.e., its actual diameter, is generally slightly larger than the final wafer's nominal diameter. In other words, the actual diameter D1 of wafer 10 is slightly larger than its nominal diameter D2. If the actual diameter D1 of wafer 10 is 300mm, then its corresponding pixel value X1 is 300 / 0.3222≈931 pixels. Since pixel value X1 is greater than pixel value X2, it can be concluded that the nominal diameter D2 of wafer 10 falls within the outline of the imaging image 20. That is, the imaging outline 30 of wafer 10 in imaging image 20 is completely within the outer edge outline range of imaging image 20, meaning the camera can completely illuminate wafer 10.

[0061] If the imaging outline 30 of wafer 10 is within the outline of imaging image 20, then the area outside the imaging outline 30 of wafer 10 in imaging image 20 is cropped; thus, in order to ensure that the imaging outline 30 of wafer 10 does not include the background pattern on the table when judging the defect level, the area outside the imaging outline 30 of wafer 10 needs to be cropped to prevent judgment errors.

[0062] If the actual diameter D1 of wafer 10 is smaller than its specified diameter D2, then the calculated pixel value X1 is smaller than the pixel value X2. Therefore, it can be concluded that the specified diameter D2 of wafer 10 falls outside the outline of the image 20. In other words, the image outline 30 of wafer 10 in the image 20 is completely outside the outer edge of the image 20. This means the camera cannot completely illuminate wafer 10, and when determining the defect level, the image outline 30 of wafer 10 will be considered to include the background pattern on the table, resulting in an incorrect judgment. At this point, the camera's focus is readjusted, and the image is re-developed to ensure that the actual diameter D1 of wafer 10 is greater than its specified diameter D2. This process continues until the calculated pixel value X1 corresponding to the actual diameter D1 is greater than the pixel value X2 corresponding to the specified diameter D2, before proceeding to the next step of the judgment.

[0063] S2. Then, divide the actual outline of the wafer 10 into several regions; then, based on the projection shape of each defect in the imaging outline 30 of the wafer 10 in the imaging image 20 and its quantity in different regions, determine the defect level of the wafer 10.

[0064] Specifically, since the actual outline of wafer 10, i.e., its actual diameter D1, is greater than its nominal diameter D2, a circle with half the nominal diameter D2 as its diameter is designated as the inner circle, and a circle with nominal diameter D2 as its outer circle, with both the inner and outer circles concentrically positioned with respect to wafer 10. The actual outline of wafer 10 is divided into three regions by the inner and outer circles: the area occupied by the inner circle is the inner ring region 13; the annular region between the inner and outer circles is the outer ring region 12; and the annular region between the outer circle and the actual outline of wafer 10 is the outer edge region 11. The diameter of the inner circle is half the diameter of the outer circle, which is also the nominal diameter D2.

[0065] Furthermore, the shapes of defects within the imaging contour 30 corresponding to wafer 10 mainly fall into three categories, including, such as Figure 2 As shown, there are three defects: a long, arc-shaped structure resembling a flame tail (defect 1, 40), a ring-shaped structure with uneven width (defect 2, 50), and an elliptical structure (defect 3, 60). Defects 1, 40, 2, 50, and 3, 60 do not overlap.

[0066] Since the intensity of photoluminescence in the tester is proportional to the concentration of non-equilibrium minority carriers at the defect location, and defects are strong recombination centers for minority carriers, the reduced minority carrier concentration in this region leads to a weakened fluorescence effect, which appears as dark dots, lines, or certain areas on the image. In other words, in defective areas, some defect shapes appear black when the color is severe, followed by gray. For defect-free areas within the wafer, where minority carrier recombination is less frequent, the non-defective parts appear brighter on the image. Therefore, when determining the defect level of wafer 10, it is also necessary to consider the development color of each defect within the imaging contour 30 of the imaging image 20. The development color of the defect within the imaging contour 30 includes gray and black.

[0067] Furthermore, defect 1 40 is located in the outer edge area 11; defect 2 50 is located in the inner ring area 13 and the outer ring area 12; and defect 3 60 is located in the inner ring area 13.

[0068] As the number of defect shapes decreases in the three regions, the defect level of wafer 10 increases sequentially. The criteria for determining the defect level of wafer 10 are as follows:

[0069] First, determine the number of defects 40 in the outer edge region 11. Preferably, the number of defects 40 in the outer edge region 11 is not less than two.

[0070] Secondly, determine the quantity and color of defect 2 50 in the inner ring area 13 and / or the outer ring area 12; and in the inner ring area 13 and / or the outer ring area 12, the color of defect 2 50 is selected as black and gray respectively.

[0071] Furthermore, the number of defects in the inner ring area 13 is 60.

[0072] In this embodiment, the defect level of wafer 10 is divided into eight levels, from C to A to Good, indicating that the severity of the defects gradually decreases, or the defect level gradually improves. Specifically, level C includes three grades: C3, C2, and C1, from lowest to highest; level B includes two grades: B2 and B1; level A includes two grades: A2 and A1; and the best grade is Good.

[0073] The specific determination is as follows:

[0074] C3 mode is:

[0075] If the number of defects 40 in the outer edge area 11 is ≥2, it is classified as grade C3. That is, as long as the number of defects 40 in the outer edge area 11 is not less than two, and there are no defects 50 or 60 in the outer ring area 12 and / or the inner ring area 13, it is grade C3.

[0076] For example: when the entire imaging contour 30 is clearly black, and the number of defects 40 in the outer edge region 11 is ≥2; or

[0077] A black color appears at the center of the inner ring area 13, and defect 250 appears in either the outer ring area 12 or the inner ring area 13, and the number of defect 40 in the outer edge area 11 is ≥2; or

[0078] Black or light black areas are present in both the outer ring area 12 and / or the inner ring area 13, and the number of defects 40 in the outer edge area 11 is ≥2, etc.

[0079] C2 is:

[0080] When the number of defects 40 in the outer edge region 11 is less than 2

[0081] Defect 250 appears in both outer ring area 12 and inner ring area 13. The rings in defect 250 are not evenly spaced and are black in both outer ring area 12 and inner ring area 13.

[0082] C1 mode is:

[0083] When defect 2 50 appears in both outer ring area 12 and inner ring area 13, the rings in defect 2 50 are not evenly spaced, and the color in both outer ring area 12 and inner ring area 13 is gray.

[0084] B2 is:

[0085] When there is defect 2 50 or defect 3 60 in the inner ring area 13, and there is no defect 2 50 in the outer ring area 12, the rings in defect 2 50 are not evenly spaced, and the color in the outer ring area 12 or the inner ring area 13 is black.

[0086] B1 is:

[0087] When there is defect 2 50 or defect 3 60 in the inner ring area 13, and there is no defect 2 50 in the outer ring area 12, the rings in defect 2 50 are not evenly spaced, and the color in the outer ring area 12 or the inner ring area 13 is gray.

[0088] A2 level is:

[0089] When the inner ring area 13 has no defects 2 50 and 3 60, and the outer ring area 12 has defect 2 50, the rings in defect 2 50 are not evenly spaced, and the color in the outer ring area 12 is black.

[0090] A1 is:

[0091] When the inner ring area 13 has no defects 2 50 and 3 60, and the outer ring area 12 has defect 2 50, the rings in defect 2 50 are not evenly spaced, and the color in the outer ring area 12 is gray.

[0092] The "Good" grade is:

[0093] When there is a defect 50 in the inner ring area 13 and / or the outer ring area 12, the rings in the defect 50 are evenly spaced, and the color in the inner ring area 13 and / or the outer ring area 12 is gray; or

[0094] When there is an incomplete annular defect 50 in the inner ring area 13 and / or the outer ring area 12, the lines in the defect 50 are distributed intermittently, and the color in the inner ring area 13 and / or the outer ring area 12 is gray; or

[0095] No defects 250 and 60 are found in the inner ring area 13 and outer ring area 12. The wafer surface defect level determination method designed in this invention can accurately calculate the true area of ​​the tested sample and quickly locate the region of the tested sample; then, based on the defect shape and the color of the imaging position, the location of different defects is determined, and the defect level of the wafer is comprehensively determined, so as to provide accurate defect analysis for production.

[0096] The embodiments of the present invention have been described in detail above. These descriptions are merely preferred embodiments and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the present invention should still fall within the patent coverage of the present invention.

Claims

1. A method for determining the surface defect level of a wafer, characterized in that the steps include... include: An image is taken of the wafer under test, and the pixel value corresponding to the actual diameter of the wafer is compared with the pixel value corresponding to the specified diameter of the wafer. A pixel reference coefficient is obtained based on the ratio of the outer edge diameter of the image to its pixel value. Based on the pixel reference coefficient, the pixel values ​​corresponding to the actual diameter and the specified diameter are obtained respectively. The pixel values ​​corresponding to the actual diameter and the specified diameter are then compared to determine whether the image outline of the wafer is within the outline of the developed image of the wafer. If the imaging contours of the wafers are all within the contours of the imaging image, then the area outside the imaging contours corresponding to the wafers in the imaging image is cropped. Then, divide the actual outline of the wafer into several regions; The defect level of the wafer is determined based on the projection shape of each defect within the imaging profile corresponding to the wafer in the imaging image, the number of defects in different regions, and the development color of the defects in the imaging image.

2. The method for determining the surface defect level of a wafer according to claim 1, characterized in that, When the pixel value corresponding to the actual diameter is less than the pixel value corresponding to the specified diameter; Therefore, the imaging contour of the wafer is outside the imaging image contour; Adjust the camera focal length and re-develop the image until the pixel value corresponding to the actual diameter is greater than the pixel value corresponding to the specified diameter.

3. The method for determining the surface defect level of a wafer according to any one of claims 1-2, characterized in that, The step of further dividing the actual outline of the wafer into several regions includes: The wafer is divided into concentric inner and outer circles within its actual outline. The area occupied by the inner circle is the inner ring area; The annular region between the inner circle and the outer circle is the outer ring region; and The annular region between the outer circle and the actual contour of the wafer is the outer edge region.

4. The method for determining the surface defect level of a wafer according to claim 3, characterized in that, The inner circle diameter is half the outer circle diameter, and the outer circle diameter is the specified diameter.

5. The method for determining the surface defect level of a wafer according to claim 4, characterized in that, The specified diameter is smaller than the actual diameter of the wafer.

6. A method for determining the surface defect level of a wafer according to claim 4 or 5, characterized in that, The step of determining the defect level of the wafer further includes determining the development color within the imaging outline corresponding to the wafer in the imaging image based on each type of defect; The defects are rendered in the image in shades of gray and black. As the number of defects in the region decreases, the defect level of the wafer increases sequentially.

7. The method for determining the surface defect level of a wafer according to claim 6, characterized in that, The shape of the defect within the imaging profile corresponding to the wafer includes: One defect of the elongated, arc-shaped structure; The second defect is a ring-shaped structure with uneven width; and The third defect of the elliptical structure; The first defect, the second defect, and the third defect do not overlap.

8. The method for determining the surface defect level of a wafer according to claim 7, characterized in that, The defect is distributed in the outer edge region; The second defect is distributed in the inner ring area and the outer ring area; The three defects are distributed in the inner ring area.

9. A method for determining the surface defect level of a wafer according to claim 7 or 8, characterized in that, The criteria for determining the defect level of the wafer include, in order: The number of defects in the outer edge region; The quantity and color of defect 2 in the inner ring area and / or the outer ring area; and The number of defect three in the inner ring area; Wherein, the number of defects in the outer edge region is not less than two; The color of the second defect in the inner ring area and / or the outer ring area is selected as black and gray respectively.

Images