A method, equipment, and system for fabricating a wafer appearance defect detection formula program.

Abstract

This invention relates to the field of wafer appearance inspection formulation technology, specifically to a method, equipment, and system for creating a wafer appearance defect inspection formulation program. The method includes the following steps: Step S1: Loading the wafer and confirming its information; Step S2: Setting wafer angle correction to level the wafer; Step S3: Wafer offset correction to align the wafer center with the motion platform; Step S4: Setting a grain distribution map, matching the theoretical distribution map with the actual distribution map in terms of position; Step S5: Determining the detection elements, randomly selecting a region from the grain distribution map as the detection element, and sequentially planning and selecting the remaining regions to determine multiple detection elements; Step S6: Sequentially acquiring photographs of the detection elements; Step S7: Detection training to form a wafer appearance defect inspection formulation program.

Description

Technical Field

[0001] This invention relates to the field of wafer appearance inspection formulation technology, and more specifically, to a method, equipment, and system for creating a wafer appearance defect inspection formulation program. Background Technology

[0002] A wafer is a silicon wafer used to fabricate silicon semiconductor circuits; its raw material is silicon. High-purity polycrystalline silicon is dissolved, doped with silicon crystal seeds, and then slowly pulled out to form a cylindrical single-crystal silicon ingot. After grinding, polishing, and slicing, the silicon crystal ingot is formed into a silicon wafer, or crystal.

[0003] Before wafers are put into use, they usually need to be inspected for appearance defects to eliminate defective products. Since the die distribution of different wafer models is different, the existing wafer appearance defect inspection equipment needs to be used to inspect and identify the formula for different wafer models. The manufacturing process is cumbersome and therefore inconvenient to use. Summary of the Invention

[0004] This invention provides a method for preparing a wafer appearance defect detection formula program, which can overcome some or all of the defects of the prior art.

[0005] According to a method for creating a wafer appearance defect detection recipe program based on the present invention, which is used to create recipes for different types of wafers in software, the method specifically includes the following steps.

[0006] Step S1: Load the wafer and confirm the wafer information;

[0007] Step S2: Set wafer angle correction to flatten the wafer;

[0008] Step S3: Wafer offset correction to align the wafer center with the motion platform;

[0009] Step S4: Set the grain distribution map and match the theoretical distribution map with the actual distribution map in terms of position;

[0010] Step S5: Determine the detection element. Select any region in the grain distribution map as the detection element, and then sequentially plan and select the remaining regions to determine multiple detection elements.

[0011] Step S6: Collect images of the detected elements sequentially; such as Figure 8 As shown;

[0012] Step S7: Detection training to form a wafer appearance defect detection formula procedure.

[0013] Preferably, the wafer information in step S1 specifically includes the product name, outer diameter, and notch type.

[0014] The outer diameter determines the detection range of the entire formula, while the name and notch type determine the formula of this type of product.

[0015] Preferably, step S2 specifically includes the following steps:

[0016] Step S21: Set the objective lens to the required magnification according to the detection resolution requirements;

[0017] The preferred magnifications include five groups: 2.5X (7.2μm / pixel), 5X (3.6μm / pixel), 7.5X (2.4μm / pixel), 10X (1.8μm / pixel), and 20X (0.9μm / pixel). Adjusting the objective lens magnification can better ensure detection resolution and...

[0018] Step S22: Set the lighting parameters for bright and dark fields. The default values ​​for dark field lighting and bright field lighting are 150, which are mutually exclusive. The values ​​can be modified as needed.

[0019] The bright field section includes six film options: Red, Blue, Green, Baseline, Brighter, and Brightest (the film options are available via an interface, but not hardware-based).

[0020] Step S23: Set the focus parameters by adjusting the lens by raising and lowering it along the z-axis. After adjustment, input the focus height and set the height deviation.

[0021] Preferably, step S23 can better ensure that the subsequently acquired images are clear.

[0022] Step S24, setting motion platform controller parameters: XY direction includes inching and continuous motion, inching step distance is the distance moved in each step and stops after movement, Z axis direction is manually used to adjust the convergence surface; continuous motion is continuous movement without stopping;

[0023] Specifically, the preferred step distance is 60, where 60 represents a movement of 60mm per cycle. The XY axis allows control of the camera's field of view movement in eight directions: up, down, left, and right. The center icon indicates the camera's field of view has been moved to the center position. The Z-axis allows manual adjustment of the convergence plane, with adjustments made in micrometer increments (note that the unit is mm). Continuous motion indicates continuous movement of the motion mechanism, which is started by pressing and holding the left mouse button and stopped by releasing it.

[0024] Step S25, Autofocus: Adjust the XY direction of the lens through the motion platform so that the center of the lens is aligned with the center of the wafer (x0, y0);

[0025] Specifically, the motion platform carries the wafer and moves it to adjust the relative position between the lens and the wafer in the XY direction to adjust the field of view, thereby aligning the center of the lens with the center of the wafer. This method is preferred because the wafer's position will continue to be adjusted in subsequent steps. Therefore, any positional deviation of the wafer during the movement of the motion platform can still be compensated for later. However, since parameters such as lens magnification and focus height have already been set, any lens position shift during movement will require readjustment of these parameters to ensure clear image acquisition.

[0026] Step S26, Mark as left template: Adjust the XY direction of the lens using the motion platform to find the marked feature on the left side of the wafer and set it as the template. Use the Hough transform to identify the circle method to find the coordinates of the center point (x, y). l y l ),like Figure 2 As shown;

[0027] Step S27, Mark as right-side template: Adjust the lens xy direction using the motion platform to find the marking feature on the right side of the wafer and set it as the template. Use the Hough transform to identify the circle method to calculate the center point coordinates (x, y). r y r );

[0028] Step S28. Calculate the wafer mounting angles in the x and y directions. and In the formula, (x0, y0) represents the original distance between the left and right marking features of the wafer in the XY direction.

[0029] Specifically, for steps S26-S28, if the wafer is not placed flat, the distance between the center point of the left template and the center point of the right template in the image acquired from directly above (orthogonally decomposed into the X and Y directions) is the projection of the original distance onto the XY direction (i.e., the horizontal plane). Therefore, the wafer's mounting angle in the x and y directions can be obtained by converting the projected distance to the original distance. It is understandable that the original distance is a control item in manufacturing, and the size is precise. The wafer being misaligned will cause the image recognition size to be smaller than the actual size.

[0030] Step S29: Adjust the motion platform controller according to the wafer mounting angle to ensure the wafer is placed flat.

[0031] Preferably, step S3 specifically includes the following steps:

[0032] Step S31: Select wafer marker features and set them as templates; use the Hough transform method to identify circles and calculate the center point coordinates (X, Y); Figure 3 As shown;

[0033] Step S32, calculate the offsets in the x and y directions: Δx = X - x0, Δy = Y - y0;

[0034] Step S33: The motion platform is corrected according to the calculated offset to ensure that the centers of the lens, the wafer and the motion platform are on a straight line.

[0035] Specifically, the preferred method is to place the wafer on the motion platform using jaws. The jaws can directly correct the wafer based on the offset to ensure that the center of the lens, the wafer, and the motion platform are collinear, thereby enabling better subsequent image acquisition and scanning.

[0036] Preferably, step S4 specifically includes the following steps:

[0037] Step S41: Start the equipment, scan the entire wafer and obtain the measured image of the grains;

[0038] Step S42: Import the designed distribution map file to determine the areas that need to be detected; such as... Figure 4 As shown;

[0039] Specifically, the distribution map file in step S42 is a file that specifies which areas (mainly the grains, which are specified according to the design drawing) in the wafer need to be inspected and which areas do not need to be inspected.

[0040] Step S43: Set the grain width and height alignment in the distribution map file, and fine-tune the grain width and height according to the grain size on the measured map.

[0041] Step S43, as a preferred option, can better match the processing error of the actual grains;

[0042] Step S44, EBR settings: Edge spherical removal. (e.g.) Figure 5 As shown;

[0043] Specifically, the EBR setting can control and filter the effective grain distribution using the actual given inner diameter size.

[0044] Step S45, as Figure 6 As shown; use the mouse wheel to zoom in on the grain distribution map in the alignment window, move the view to the effective grain positions at the top and bottom of the wafer, select the theoretical cell with the mouse, and align the top and bottom edges;

[0045] Step S46, as follows Figure 7 As shown; zoom in on the grain distribution map using the mouse wheel, move the view to the leftmost and rightmost effective grain positions on the wafer, align them with the leftmost and rightmost effective grains in the scanned grain distribution map, and complete the setting of the grain distribution map.

[0046] Specifically, steps S45 and S46 enable the grain distribution map to match the positions of the previously scanned grain measurement map to facilitate subsequent matching and identification detection.

[0047] Preferably, in step S5, each detection element is designated as region 1, region 2, and other regions.

[0048] Preferably, step S7 specifically includes the following steps:

[0049] Step S71, set training parameters

[0050] Step S72, Training: After setting up, start training to generate the gold template.

[0051] As a preferred option, such as Figure 9 As shown; the training parameters in step S71 include the number and distribution of training chips. The system defaults to 23 random chips. Training chips can be manually added or deleted. The number of training chips is 9 ≤ number of training chips ≤ 30. The number of training times is provided in three ways: OnlyOnce, Every Wafer, and Every Lot.

[0052] Combination Figure 10 The gold template in step S72 uses a median filtering algorithm, and the testers have checked and confirmed that there are no abnormalities such as misalignment or ghosting.

[0053] The present invention also provides a wafer appearance defect detection recipe program fabrication device, including a motion platform, a lens, a memory and a processor, which is used to implement the steps of the aforementioned wafer appearance defect detection recipe program fabrication method.

[0054] The present invention also provides a wafer appearance defect detection recipe preparation system, characterized in that: it includes an input device for inputting parameters and an output device for outputting results, as well as the wafer appearance defect detection recipe preparation equipment. Attached Figure Description

[0055] Figure 1 This is a flowchart of a wafer appearance defect detection formula preparation method in Example 1;

[0056] Figure 2 This is a schematic diagram illustrating step S26 in Example 1;

[0057] Figure 3 This is a schematic diagram illustrating step S31 in Example 1;

[0058] Figure 4 This is a schematic diagram illustrating step S42 in Example 1;

[0059] Figure 5This is a schematic diagram illustrating step S44 in Example 1;

[0060] Figure 6 This is a schematic diagram illustrating step S45 in Example 1;

[0061] Figure 7 This is a schematic diagram illustrating step S46 in Example 1;

[0062] Figure 8 This is a schematic diagram illustrating step S6 in Example 1;

[0063] Figure 9 This is a schematic diagram illustrating step S71 in Example 1;

[0064] Figure 10 This is a schematic diagram illustrating step S72 in Example 1. Detailed Implementation

[0065] To further understand the content of this invention, the invention will be described in detail with reference to the embodiments. It should be understood that the embodiments are merely illustrative and not limiting of the invention.

[0066] Example 1

[0067] Combination Figure 1-10 This embodiment provides a method for creating a wafer appearance defect detection recipe program, which is used to create recipes for different types of wafers on software. The method specifically includes the following steps:

[0068] Step S1: Load the wafer and confirm the wafer information;

[0069] Step S2: Set wafer angle correction to flatten the wafer;

[0070] Step S3: Wafer offset correction to align the wafer center with the motion platform;

[0071] Step S4: Set the grain distribution map and match the theoretical distribution map with the actual distribution map in terms of position;

[0072] Step S5: Determine the detection element. Select any region in the grain distribution map as the detection element, and then sequentially plan and select the remaining regions to determine multiple detection elements.

[0073] Step S6: Collect images of the detected elements sequentially; such as Figure 8 As shown;

[0074] Step S7: Detection training to form a wafer appearance defect detection formula procedure.

[0075] Specifically, this embodiment can quickly and conveniently train a wafer appearance defect detection formula program through the above method, which can then be better used for subsequent wafer inspection of different models.

[0076] In this embodiment, the wafer information in step S1 specifically includes product name, outer diameter, and notch type.

[0077] The outer diameter determines the detection range of the entire formula, while the name and notch type determine the formula of this type of product.

[0078] In this embodiment, step S2 specifically includes the following steps:

[0079] Step S21: Set the objective lens to the required magnification according to the detection resolution requirements;

[0080] The required magnification in this embodiment includes five sets: 2.5X (7.2μm / pixel), 5X (3.6μm / pixel), 7.5X (2.4μm / pixel), 10X (1.8μm / pixel), and 20X (0.9μm / pixel). Adjusting the objective lens magnification can better ensure the detection resolution.

[0081] Step S22: Set the lighting parameters for bright and dark fields. The default values ​​for dark field lighting and bright field lighting are 150, which are mutually exclusive. The values ​​can be modified as needed.

[0082] The bright field includes six film options: Red, Blue, Green, Baseline, Brighter, and Brightest (the film options are available via an interface, but not hardware).

[0083] Step S23: Set the focus parameters by adjusting the lens by raising and lowering it along the z-axis. After adjustment, input the focus height and set the height deviation.

[0084] In this embodiment, step S23 can better ensure that the subsequently acquired images are clear.

[0085] Step S24, setting motion platform controller parameters: XY direction includes inching and continuous motion, inching step distance is the distance moved in each step and stops after movement, Z axis direction is manually used to adjust the convergence surface; continuous motion is continuous movement without stopping;

[0086] Specifically, in this embodiment, the step distance is 60, which represents a movement of 60mm per cycle. The XY axis can control the movement of the camera's field of view in eight directions: up, down, left, and right. The center icon indicates the position of the camera's field of view. The Z-axis is manually adjusted to adjust the convergence plane, and the adjustment is done in micrometer increments (note that the unit is mm). Continuous motion indicates continuous movement of the motion mechanism, which is started by pressing and holding the left mouse button and stopped by releasing it.

[0087] Step S25, Autofocus: Adjust the XY direction of the lens through the motion platform so that the center of the lens is aligned with the center of the wafer (x0, y0);

[0088] Specifically, the motion platform carries the wafer to adjust the relative position between the lens and the wafer in the XY direction to adjust the field of view, thereby aligning the lens center with the wafer center. The purpose of this method in this embodiment is that the wafer's position will continue to be adjusted in subsequent steps. Therefore, when the motion platform carries the wafer, any positional deviation of the wafer can still be compensated for later. However, the lens magnification, focus height, and other parameters have already been set. If the lens is moved further, any lens position shift during the movement will require further adjustment of the aforementioned parameters to ensure clear image acquisition.

[0089] Step S26, Mark as left template: Adjust the XY direction of the lens using the motion platform to find the marked feature on the left side of the wafer and set it as the template. Use the Hough transform to identify the circle method to find the coordinates of the center point (x, y). l y l ),like Figure 2 As shown;

[0090] Step S27, Mark as right-side template: Adjust the lens xy direction using the motion platform to find the marking feature on the right side of the wafer and set it as the template. Use the Hough transform to identify the circle method to calculate the center point coordinates (x, y). r y r );

[0091] Step S28. Calculate the wafer mounting angles in the x and y directions. and In the formula, (x0, y0) represents the original distance between the left and right marking features of the wafer in the XY direction.

[0092] Specifically, for steps S26-S28, if the wafer is not placed flat, the distance between the center point of the left template and the center point of the right template in the image acquired from directly above (orthogonally decomposed into the X and Y directions) is the projection of the original distance onto the XY direction (i.e., the horizontal plane). Therefore, the wafer's mounting angle in the x and y directions can be obtained by converting the projected distance to the original distance. It is understandable that the original distance is a control item in manufacturing, and the size is precise. The wafer being misaligned will cause the image recognition size to be smaller than the actual size.

[0093] Step S29: Adjust the motion platform controller according to the wafer mounting angle to ensure the wafer is placed flat.

[0094] In this embodiment, step S3 specifically includes the following steps:

[0095] Step S31: Select wafer marker features and set them as templates; use the Hough transform method to identify circles and calculate the center point coordinates (X, Y); Figure 3 As shown;

[0096] Step S32, calculate the offsets in the x and y directions: Δx = X - x0, Δy = Y - y0;

[0097] Step S33: The motion platform is corrected according to the calculated offset to ensure that the centers of the lens, the wafer and the motion platform are on a straight line.

[0098] Specifically, in this embodiment, the wafer is placed on the motion platform by a chuck. The chuck can directly correct the wafer according to the offset to ensure that the center of the lens, the wafer and the motion platform are collinear; thus enabling better subsequent image acquisition and scanning.

[0099] In this embodiment, step S4 specifically includes the following steps:

[0100] Step S41: Start the equipment, scan the entire wafer and obtain the measured image of the grains;

[0101] Step S42: Import the designed distribution map file to determine the areas that need to be detected; such as... Figure 4 As shown;

[0102] Specifically, the distribution map file in step S42 is a file that specifies which areas (mainly the grains, which are specified according to the design drawing) in the wafer need to be inspected and which areas do not need to be inspected.

[0103] Step S43: Set the grain width and height alignment in the distribution map file, and fine-tune the grain width and height according to the grain size on the measured map.

[0104] Step S43 in this embodiment can better match the processing error of the actual grain;

[0105] Step S44, EBR settings: Edge spherical removal. (e.g.) Figure 5 As shown;

[0106] Specifically, the EBR setting can control and filter the effective grain distribution using the actual given inner diameter size.

[0107] Step S45, as Figure 6 As shown; use the mouse wheel to zoom in on the grain distribution map in the alignment window, move the view to the effective grain positions at the top and bottom of the wafer, select the theoretical cell with the mouse, and align the top and bottom edges;

[0108] Step S46, as follows Figure 7 As shown; zoom in on the grain distribution map using the mouse wheel, move the view to the leftmost and rightmost effective grain positions on the wafer, align them with the leftmost and rightmost effective grains in the scanned grain distribution map, and complete the setting of the grain distribution map.

[0109] Specifically, steps S45 and S46 enable the grain distribution map to match the positions of the previously scanned grain measurement map to facilitate subsequent matching and identification detection.

[0110] In this embodiment, in step S5, each detection element is defined as region 1, region 2, and other regions.

[0111] In this embodiment, step S7 specifically includes the following steps:

[0112] Step S71, set training parameters

[0113] Step S72, Training: After setting up, start training to generate the gold template.

[0114] In this embodiment, as Figure 9 As shown; the training parameters in step S71 include the number and distribution of training dies. The system defaults to 23 random dies. Training dies can be manually added or deleted. The number of training dies is 9 ≤ number of training dies ≤ 30. The number of training times is provided in three ways: Only Once, Every Wafer, and Every Lot.

[0115] Combination Figure 10 The gold template in step S72 uses a median filtering algorithm, and the testers have checked and confirmed that there are no abnormalities such as misalignment or ghosting.

[0116] This embodiment also provides a wafer appearance defect detection recipe program fabrication device, including a motion platform, a lens, a memory, and a processor, which is used to implement the steps of the aforementioned wafer appearance defect detection recipe program fabrication method.

[0117] This embodiment also provides a wafer appearance defect detection recipe program creation system, characterized in that: it includes an input device for inputting parameters and an output device for outputting results, as well as the wafer appearance defect detection recipe program creation equipment.

[0118] It is readily understood that those skilled in the art can combine, split, or reorganize the embodiments provided in this application to obtain other embodiments, all of which do not exceed the protection scope of this application.

[0119] The present invention and its embodiments have been described above illustratively. This description is not restrictive, and the embodiments shown are only part of the embodiments of the present invention. The actual structure is not limited thereto. Therefore, if those skilled in the art are inspired by this description and design similar structures and embodiments without departing from the spirit of the present invention, they should all fall within the protection scope of the present invention.

Claims

1. A method for creating a wafer appearance defect detection recipe program, used for creating recipes on software for wafers of different models, characterized in that: Specifically comprising the following steps, Step S1: wafer information is confirmed after wafer loading; Step S2: wafer angle correction is set for wafer leveling, specifically comprising: Step S21, according to the detection resolution requirement, set the objective lens to the required magnification; Step S22, set the bright-dark field illumination parameters; Step S23, set the focusing parameters, focus through the z-axis lifting lens, input the focusing height after focusing and set the height deviation; Step S24, motion platform controller parameter setting: XY direction includes inching and continuous motion, inching step distance is the size of each step movement distance and stops after moving, Z-axis direction uses manual adjustment to adjust the focusing plane; continuous motion is continuous movement without stopping; Step S25, automatic focusing: adjust the lens XY direction through the motion platform to align the lens center with the wafer center (x0, y0); Step S26, mark the left template: adjust the lens XY direction through the motion platform, find the wafer left mark feature, and set as a template, use the hough transform to identify the circle method to obtain the center point coordinates (x l , y l ) Step S27, Mark as right-side template: Adjust the lens xy direction using the motion platform to find the marking feature on the right side of the wafer and set it as the template. Use the Hough transform to identify the circle method to find the center point coordinates (x, y). r y r ); Step S28, find the wafer corner angle in x and y directions and where (x0, y0) is the original distance between the left and right mark features in XY direction. Step S29, adjust the motion platform controller according to the wafer angle to ensure wafer leveling; Step S3: wafer offset correction is set to keep the wafer center consistent with the motion platform, specifically comprising: Step S31, select the wafer mark feature and set it as a template, use the hough transform to identify the center point coordinates (X, Y); Step S32, calculating the xy-direction offset , ; Step S33, the motion platform is corrected according to the offset to ensure that the centers of the lens, wafer and motion platform are on a straight line; Step S4: set the die distribution map to match the theoretical distribution map with the actual distribution map in position; Step S5: determine the detection elements, select any area in a die distribution map as a detection element, and sequentially plan to select the remaining areas to determine multiple detection elements; Step S6: sequentially collect detection element photos; Step S7: detection training to form a wafer appearance defect detection recipe program.

2. The wafer appearance defect detection recipe program making method according to claim 1, characterized by: The wafer information of step S1 specifically includes product name, outer diameter and notch type.

3. The method of claim 1, wherein the method further comprises: Step S4 specifically comprises the following steps, ​ Step S41, start the device, scan the entire wafer and obtain the die measurement map; Step S42, import the designed distribution map file to determine the area to be detected; Step S43, set the die width and height alignment in the distribution map file, and fine-tune the die width and height according to the die size on the measurement map; Step S44, EBR setting: edge ball removal; Step S45, zoom in on the die distribution map in the alignment window through the mouse wheel, move the field of view to the top and bottom effective die positions of the wafer, select the theoretical cell with the mouse, and align the top and bottom edges; Step S46, zoom in on the die distribution map through the mouse wheel, move the field of view to the left and right effective die positions of the wafer, and align them with the left and right effective dies of the scanned die distribution map to complete the setting of the die distribution map.

4. The method of claim 1, wherein the method further comprises: In step S5, each detection element is defined as No. 1 area, No. 2 area and other areas. ​ 5. The method of claim 1, wherein the method further comprises: Step S7 specifically comprises the following steps, ​ Step S71, set the training parameters Step S72, training: start training after setting is completed, and generate a golden template.

6. The method of claim 5, wherein the method further comprises: The training parameters in step S71 include the number and distribution of training grains, and the training times provide three options of Only Once / Every Wafer / Every Lot for selection; ​ The golden template in step S72 is a median filter algorithm, and no misplacement or ghosting is confirmed by the tester.

7. A wafer appearance defect detection recipe program making apparatus comprising a motion platform, a lens, a memory and a processor, characterized by: The steps for implementing the wafer appearance defect detection recipe program making method of any one of claims 1-6.

8. A wafer appearance defect detection recipe program making system characterized by comprising: The wafer appearance defect detection recipe program making device comprises input devices for inputting parameters, output devices for outputting results, and the wafer appearance defect detection recipe program making device of claim 7.

Images