A method for detecting corner defects of semiconductor masks

Abstract

The present invention discloses a method for detecting corner defects in semiconductor masks. The method comprises the following steps: S1: obtaining an actual mask image and a rendering of the corresponding mask design; S2: marking the corresponding corner positions on the rendering according to the mask design; S3: registering the rendering with the actual image; S4: performing the following operations on the rendering and the actual image at each corner position: locating sub-pixel edges and connecting the located sub-pixel edge points; S5: obtaining corner features from the connected sub-pixel edge points; S6: setting a threshold for the corresponding corner feature and determining whether the corner is a defect based on the threshold. This method can improve the noise sensitivity of corner detection, reduce corner pseudo-defects, and improve corner detection accuracy.

Description

Technical Field

[0001] The invention relates to a method for detecting corner defects of semiconductor masks, belonging to the technical field of semiconductor mask defect detection. Background Art

[0002] Currently, semiconductor mask defect detection typically uses a DIE to DB (D2DB) inspection model, comparing images of the final mask with the mask design files to identify defects. While this method is relatively easy to detect for typical defects like CDs, pinholes, and holes, corner defects, due to their location at corners, often contain more information but also introduce more noise. This makes corner defects more difficult to detect and more prone to false defects. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to overcome the defects of the prior art and provide a method for detecting corner defects of semiconductor masks, by which the sensitivity of corner detection to noise can be improved, corner pseudo defects can be reduced, and the accuracy of corner detection can be improved.

[0004] In order to solve the above technical problems, the technical solution of the present invention is: a method for detecting corner defects of semiconductor masks, the method comprising the following steps:

[0005] S1: Obtain the actual image of the mask and the corresponding rendering of the mask design;

[0006] S2: Mark the corresponding corner positions on the rendering according to the mask design drawing;

[0007] S3: align the rendered image with the actual image;

[0008] S4: performing the following operations on the rendering image and the actual image at each corner position: locating the sub-pixel edge and connecting the located sub-pixel edge points;

[0009] S5: Obtain corner features for connected sub-pixel edge points;

[0010] S6: Setting a threshold value of the corresponding corner feature, and determining whether the corner is a defect based on the threshold value.

[0011] Furthermore, in step S1, the actual image of the mask is collected by AOI inspection equipment.

[0012] Furthermore, the sub-pixel edge positioning and the connection of the positioned sub-pixel edge points are specifically as follows:

[0013] Calculate the gradient of the image; wherein the gradient includes the gradient in the horizontal direction and the vertical direction of the image;

[0014] Determine sub-pixel edge points using maximum suppression and interpolation methods based on gradient information;

[0015] Connect the determined sub-pixel edge points into an ordered sequence.

[0016] Furthermore, the corner feature includes the curvature of the corner or the curvature and the opening direction.

[0017] Furthermore, the steps to obtain the curvature of the corner are as follows:

[0018] The sub-pixel edge points to be connected are used as a data set to perform least squares circle fitting to obtain the radius r of the fitting circle;

[0019] The curvature of the corner is obtained from the radius r of the fitted circle, k=1 / r.

[0020] Furthermore, the steps to obtain the curvature and opening direction of the corner are as follows:

[0021] The sub-pixel edge points to be connected are used as a data set to perform least squares circle fitting to obtain the radius r and center of the fitting circle;

[0022] The curvature k=1 / r of the corner is obtained according to the radius r of the fitted circle. The opening direction of the corner is obtained by forming corresponding vectors by connecting each sub-pixel edge point and the center of the circle, and then summing and averaging the vectors.

[0023] Furthermore, in step S6, a threshold corresponding to the corner curvature is set according to the photolithography process level of the corner.

[0024] Furthermore, in step S6, whether the corner is a defect is determined based on the threshold value as follows:

[0025] Difference is calculated between the corner features of the corresponding corner positions of the rendering image and the actual image to obtain the difference value;

[0026] Compare the absolute value of the difference with the threshold;

[0027] Whether the corresponding corner is a defect is determined based on the comparison result of the absolute value of the difference and the threshold.

[0028] The determination of whether the corresponding corner is a defect based on the comparison result of the absolute value of the difference and the threshold is specifically as follows: when the absolute value of the difference is less than the threshold, the corresponding corner is determined not to be a defect; otherwise, it is a defect.

[0029] After adopting the above technical solution, the method proposed in the present invention can still correctly perform corner defect detection when the alignment result is not very ideal, without causing too many false defects; furthermore, this method can adjust different curvature radius thresholds according to different lithography process levels to achieve customized corner defects without causing the problem of adding too many defects or missing defects. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] FIG1( a ) is a schematic diagram of a rendering of a certain portion of a mask design diagram according to the present invention;

[0031] FIG1( b ) is a schematic diagram of an actual sampling of a certain portion of the mask corresponding to FIG1( a ) in the present invention;

[0032] FIG1( c ) is an enlarged view of a corner in FIG1( a ) of the present invention;

[0033] FIG1(d) is an enlarged view of the corner corresponding to FIG1(c) in FIG1(b) of the present invention;

[0034] Figure 2 This is the overall result diagram of the sub-pixel edge positioning connection in FIG1(b) of the present invention;

[0035] FIG3( a ) is a diagram showing the result of sub-pixel edge location connection at a corner of FIG1( c ) according to the present invention;

[0036] FIG3( b ) is a diagram showing the result of corner sub-pixel edge location connection corresponding to FIG1( d ) in the present invention; DETAILED DESCRIPTION

[0037] In order to make the contents of the present invention more clearly understood, the present invention is further described in detail below based on specific embodiments in conjunction with the accompanying drawings.

[0038] As shown in Figures 1 to 3, a method for detecting corner defects of a semiconductor mask includes the following steps:

[0039] S1: The actual image of the mask is obtained through the inspection camera in the AOI inspection equipment; and the corresponding rendering image is rendered according to the corresponding mask design drawing and the corresponding rendering model, as shown in Figure 1(a);

[0040] S2: Mark the corresponding corner positions on the rendering according to the mask design drawing;

[0041] S3: align the rendered image with the actual image;

[0042] S4: performing the following operations on the rendering image and the actual image at each corner position: locating the sub-pixel edge and connecting the located sub-pixel edge points;

[0043] S5: Obtain corner features for connected sub-pixel edge points;

[0044] S6: Setting a threshold value of the corresponding corner feature, and determining whether the corner is a defect based on the threshold value.

[0045] Specifically, the rendering model may include a simple low-pass filtering model, a scalar simulation model, or a vector simulation model according to the type of the detection mask; rendering is mainly used to make the rendered image more consistent with the captured image.

[0046] Specifically, the sub-pixel edge positioning and the connection of the positioned sub-pixel edge points are as follows:

[0047] Calculate the gradient of the image; the gradient mainly includes the gradient in the horizontal and vertical directions of the image;

[0048] Determine sub-pixel edge points using maximum suppression and interpolation methods based on gradient information;

[0049] Connect the determined sub-pixel edge points into an ordered sequence. The specific implementation method of this step can refer to the Canny edge detector or the Devernay edge detector. The specific implementation results are as follows: Figure 2 shown.

[0050] Specifically, corner features include the curvature and opening direction of the corner.

[0051] Specifically, the steps for obtaining the curvature and opening direction of the corner are as follows:

[0052] A segment of ordered sub-pixel edge points near the corner is used as a data set for least squares circle fitting to obtain the radius r and center of the fitting circle;

[0053] The curvature of the corner, k = 1 / r, is obtained based on the radius r of the fitted circle. The opening direction of the corner, i.e., the direction of the angle bisector, is obtained by forming corresponding vectors from each ordered sub-pixel edge point and the center of the circle, summing and averaging the vectors.

[0054] Specifically, in step S6, a threshold for corner curvature is set based on the lithography process level of the corner to eliminate pseudo defects with curvature differences within the acceptable range. As shown in Figure 3(a), the curvature radius of the corner on the mask design is 2.9 pixels, while the curvature radius of the corner in the actual image in Figure 3(b) is 4.7 pixels. This is a significant difference, so it can be determined that this is a corner defect.

[0055] Specifically, in step S6, whether the corner is a defect is determined based on the threshold value as follows:

[0056] Difference is calculated between the corner features of the corresponding corner positions of the rendering image and the actual image to obtain the difference value;

[0057] Compare the absolute value of the difference with the threshold;

[0058] Whether the corresponding corner is a defect is determined based on the comparison result of the absolute value of the difference and the threshold.

[0059] The determination of whether the corresponding corner is a defect based on the comparison result of the absolute value of the difference and the threshold is specifically as follows: when the absolute value of the difference is less than the threshold, the corresponding corner is determined not to be a defect; otherwise, it is a defect.

[0060] The specific embodiments described above further illustrate the technical problems, technical solutions and beneficial effects solved by the present invention. It should be understood that the above are only specific embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for detecting corner defects of semiconductor masks, characterized in that: The steps of the method include: S1: Obtain the actual image of the mask and the corresponding rendering of the mask design; S2: Mark the corresponding corner positions on the rendering according to the mask design drawing; S3: align the rendered image with the actual image; S4: performing the following operations on the rendering image and the actual image at each corner position: locating the sub-pixel edge and connecting the located sub-pixel edge points; S5: Obtain corner features for connected sub-pixel edge points; S6: Set the threshold of the corresponding corner feature and determine whether the corner is a defect based on the threshold; The corner features include the curvature or curvature of the corner and the direction of the opening; The steps to find the curvature and opening direction of the corner are as follows: The sub-pixel edge points to be connected are used as a data set to perform least squares circle fitting to obtain the radius r and center of the fitting circle; The curvature of the corner is obtained from the radius r of the fitted circle. The opening direction of the corner is obtained by forming a corresponding vector from each connected sub-pixel edge point and the center of the circle, and then summing and averaging the vectors. In step S6, whether the corner is a defect is determined based on the threshold value as follows: Difference is calculated between the corner features of the corresponding corner positions of the rendering image and the actual image to obtain the difference value; Compare the absolute value of the difference with the threshold; Whether the corresponding corner is a defect is determined based on the comparison result of the absolute value of the difference and the threshold.

2. The detection method according to claim 1, wherein In step S1, the actual image of the mask is collected by AOI inspection equipment.

3. The detection method according to claim 1, wherein The sub-pixel edge positioning and the connection of the positioned sub-pixel edge points are specifically as follows: Calculate the gradient of the image; wherein the gradient includes the gradient in the horizontal direction and the vertical direction of the image; Determine sub-pixel edge points using maximum suppression and interpolation methods based on gradient information; Connect the determined sub-pixel edge points into an ordered sequence.

4. The detection method according to claim 1, wherein In step S6, a threshold corresponding to the corner curvature is set according to the lithography process level of the corner.

5. The detection method according to claim 1, wherein The determination of whether the corresponding corner is a defect based on the comparison result of the absolute value of the difference and the threshold is specifically as follows: when the absolute value of the difference is less than the threshold, the corresponding corner is determined not to be a defect; otherwise, it is a defect.

Images