Geometric feature self-adaption-based necking detection method

Abstract

The invention discloses a necking detection method based on geometric feature self-adaption. The necking detection method comprises the following steps that S1, the longest main axis of a simulation graph is determined; s2, selecting N feature points on the longest main shaft; s3, obtaining the minimum key size of the simulation graph at each feature point; and S4, judging whether the simulation graph has a necking risk or not based on each minimum key size. By identifying the geometric centroid and the longest principal axis of a graph, adaptively selecting feature points and performing omnidirectional scanning, the measured minimum critical dimension is compared with a set safety threshold, and a risk point which is most likely to generate necking on the graph is accurately positioned, so that the automation and objectification of risk detection are realized, and the risk detection efficiency is improved. Subjectivity and experience dependence of manual marking are avoided, and consistency of detection results is guaranteed; meanwhile, the method has good adaptability to complex or deformed contours, missing detection and misjudgment are effectively avoided, and the yield and reliability of chip manufacturing are improved.

Description

Technical Field

[0001] This invention belongs to the field of semiconductors, and particularly relates to a necking detection method based on geometric feature adaptation. Background Technology

[0002] In the semiconductor industry, identifying pinching defects on contours is crucial for ensuring product yield. Especially for critical structures like hole patterns, abnormal shrinkage in the local contour can lead to increased via resistance or connection failure. Traditional inspection methods typically rely on manual experience, manually setting markers at specific locations on the target contour and performing intensive measurements of the local critical dimension (Contour CD) around these markers to detect potential pinching risks. However, this method is limited by the manual placement of markers, depends on operator subjectivity, and has poor adaptability to geometric deformations. It is prone to missed detections or misjudgments when dealing with complex shapes or contour variations, making it difficult to guarantee consistency and comprehensiveness of inspections, and ultimately failing to meet the high-precision, high-efficiency defect detection requirements of advanced process nodes.

[0003] To address the aforementioned issues, a necking detection method based on geometric feature adaptation is needed to improve the objectivity, accuracy, and reliability of the detection. Summary of the Invention

[0004] The purpose of this invention is to solve all or part of the above-mentioned problems by providing a necking detection method based on geometric feature adaptation. By identifying the geometric centroid and longest principal axis of the graphic, feature points are adaptively selected and omnidirectional scanning is performed. The measured minimum critical dimension is compared with a set safety threshold to accurately locate the risk points on the graphic most prone to necking, thereby automating and objectifying risk detection, avoiding the subjectivity and experience dependence of manual marking, and ensuring the consistency of detection results. At the same time, it has good adaptability to complex or deformed contours, effectively avoiding missed detections and misjudgments, and improving the yield and reliability of chip manufacturing.

[0005] This invention provides a necking detection method based on geometric feature adaptation, comprising the following steps: S1: confirming the longest principal axis of the simulated pattern; S2: selecting N feature points on the longest principal axis; S3: obtaining the minimum critical dimension of the simulated pattern at each of the feature points; S4: determining whether the simulated pattern has a necking risk based on each of the minimum critical dimensions. By automatically identifying the centroid and principal axis of the pattern, adaptively selecting detection points and performing omnidirectional scanning, comparing the measured minimum width with a safety threshold, the method accurately locates the necking risk, achieving automation and objectivity in detection, avoiding human subjectivity, exhibiting strong adaptability to complex contours, effectively preventing missed detections and misjudgments, and improving the yield and reliability of chip manufacturing.

[0006] The simulated graphic is obtained by simulating the design graphic through a photolithography process model after optical proximity correction; the simulated graphic is an elongated strip-shaped graphic. Verification is performed before physical manufacturing to ensure product yield and control production costs; the elongated hole-shaped graphic has a central necking region, and arranging feature points along the longest main axis can effectively cover this high-risk area, demonstrating high practicality.

[0007] The longest principal axis is determined as follows: the geometric centroid of the simulated graphic is calculated; based on the geometric centroid, the longest principal axis of the simulated graphic is determined. Determining the longest principal axis using the geometric centroid makes the determination more reliable and improves the defect detection rate. The critical dimension is the inner diameter of the simulated graphic. Aligning the detection target with electrical performance requirements enhances the practicality of risk prediction.

[0008] In step S2, the feature points are selected as follows: The length L of the longest principal axis is determined, and starting from one end of the longest principal axis, a feature point is selected every L / N+1 interval along the direction of the longest principal axis. This achieves adaptive and standardized distribution of detection points, automatically adjusting according to the actual size of different graphics, avoiding the subjectivity of manual settings, and ensuring consistent results. It effectively covers areas prone to necking in elongated hole structures, focusing the detection position on key areas of the graphic, ensuring accurate detection results, and possessing universality and reliability for various graphics.

[0009] The number of feature points is two. This approach covers the main potential risk points with minimal computational cost, avoiding increased computational load from scanning the entire spindle, thus improving detection speed and providing a practical and efficient solution while ensuring risk detection capability.

[0010] In step S3, the process of determining the minimum critical dimension is as follows: obtain the critical dimensions of the simulation graphic in all directions at the current feature point, and take the minimum value of the critical dimensions as the minimum critical dimension of the current feature point. Selecting the minimum critical dimension ensures the accuracy of risk assessment, lays a reliable foundation for setting a unified automated risk judgment, and effectively avoids missed judgments due to vague or lenient assessment standards.

[0011] The critical dimension is the inner diameter of the simulated graphic passing through the feature point. Aligning the detection target with electrical performance requirements improves the practicality of risk prediction. In step S4, the determination method is as follows: a corresponding safety threshold is set for each feature point; if the minimum critical size at each feature point is not less than its corresponding safety threshold, it is determined that there is no necking risk; otherwise, it is determined that there is a necking risk. The set safety threshold provides a unified and objective detection standard, avoids subjective differences in the judgment process, and improves detection efficiency and response speed.

[0012] The safety threshold is determined based on the design dimensions of the design graphic. The safety threshold originates from the product's design specifications and can be adaptively adjusted according to the size requirements of different technical nodes. This avoids the potential for overly lenient or overly strict decisions due to the use of fixed values, thus improving the scientific rigor of risk assessment.

[0013] The safety threshold is the design dimension minus 1-2 minimum length units. Setting a reasonable safety threshold ensures that the defect judgment criteria match the actual manufacturing level and conform to the operational standards of production reality.

[0014] Compared with existing technologies, the beneficial effects of this invention are: by identifying the geometric centroid and longest principal axis of the graphic, adaptively selecting feature points and performing omnidirectional scanning, comparing the measured minimum critical size with the set safety threshold, accurately locating the risk points on the graphic most prone to necking, realizing the automation and objectification of risk detection, avoiding the subjectivity and experience dependence of manual marking, and ensuring the consistency of detection results; at the same time, it has good adaptability to complex or deformed contours, effectively avoiding missed detections and misjudgments, and improving the yield and reliability of chip manufacturing. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the specific embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 A schematic diagram of the necking detection method provided by the prior art.

[0017] Figure 2 This is a schematic diagram of a necking detection method based on geometric feature adaptation provided by the present invention. Detailed Implementation

[0018] The technical solutions in specific embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] The existing pinning defect detection process for Hole Pattern graphics involves manually placing a marker at a fixed location on the simulated contour (e.g., at 1 / 3 of its length), and then measuring the key dimensions of the local contour within the marked area to determine if there is a risk of necking. Figure 1As shown, the solid line represents the size of the marker inspection area, while the actual inspection needs to be performed using the dashed line size. Existing methods are limited by the subjectivity of manually setting the marker position and have poor adaptability to geometric deformations, making them prone to missed detections or misjudgments in complex shapes.

[0020] Example This embodiment provides a necking detection method based on geometric feature adaptation, such as... Figure 2 As shown, it includes the following steps: S1: After the design graphic is corrected by optical proximity correction, it is simulated by photolithography process model to obtain the contour data of HolePattern simulation graphic, calculate the geometric centroid of the simulation graphic, and determine the longest principal axis of the simulation graphic. S2: Select N feature points along the longest main axis. The selection method is as follows: the length of the longest main axis is L. Starting from one end of the longest main axis, select one feature point every L / (N+1) along the direction of the longest main axis. In this embodiment, the number of feature points is set to N=2. Then, the positions of two feature points are L / 3 and 2L / 3 from the end point of the longest main axis. This can effectively cover the middle area in the long strip hole structure that is prone to necking. S3: Perform an omnidirectional scan at each feature point location, measure the critical dimensions of the simulation graphic in each direction at the feature point. The critical dimensions are the inner diameter dimensions of the simulation graphic, and select the minimum value as the minimum critical dimension corresponding to that feature point. S4: Defect Judgment: Set a safety threshold, which is set according to the design dimensions of the design graphic. Specifically, the safety threshold is set by subtracting 1-2 minimum unit lengths (DBU) from the design dimensions. DBU is the minimum length unit representing coordinates in the layout data, used to convert physical dimensions into integer coordinates for storage and calculation. The minimum critical dimensions at two feature points are compared with the corresponding safety thresholds. If the minimum critical dimensions are less than the safety threshold, it is determined that there is a risk of necking; otherwise, it is determined that there is no risk of necking. The detection process does not require manual intervention, achieving a rapid, objective, and accurate assessment of necking defects in elongated hole graphics.

[0021] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A necking detection method based on geometric feature adaptation, characterized in that, Includes the following steps: S1: Confirm the longest principal axis of the simulated graphic; S2: Select N feature points on the longest principal axis; S3: Obtain the minimum critical dimension of the simulation graphic at each of the feature points; S4: Determine whether the simulation graphic has a risk of necking based on each of the minimum critical dimensions.

2. The necking detection method based on geometric feature adaptation according to claim 1, characterized in that, The simulated graphic is a graphic obtained by simulating the design graphic through a photolithography process model after optical proximity correction; the simulated graphic is a long strip-shaped graphic.

3. The necking detection method based on geometric feature adaptation according to claim 1, characterized in that, The longest principal axis is determined by calculating the geometric centroid of the simulation graphic. Based on the geometric centroid, the longest principal axis of the simulation graphic is determined.

4. The necking detection method based on geometric feature adaptation according to claim 1, characterized in that, In step S2, the feature points are selected as follows: the length L of the longest principal axis is determined, and starting from one end of the longest principal axis, a feature point is selected every L / N+1 intervals along the direction of the longest principal axis.

5. The necking detection method based on geometric feature adaptation according to claim 1, characterized in that, The number of feature points is 2.

6. The necking detection method based on geometric feature adaptation according to claim 1, characterized in that, In step S3, the process of determining the minimum critical dimension is as follows: obtain the critical dimensions of the simulation graphic in all directions at the current feature point, and take the minimum value of the critical dimensions as the minimum critical dimension of the current feature point.

7. The necking detection method based on geometric feature adaptation according to claim 6, characterized in that, The critical dimension is the inner diameter of the simulated graphic through the feature point.

8. The necking detection method based on geometric feature adaptation according to claim 1, characterized in that, In step S4, the determination method is as follows: set a corresponding safety threshold for each feature point; if the minimum critical size at each feature point is not less than its corresponding safety threshold, it is determined that there is no risk of necking; otherwise, it is determined that there is a risk of necking.

9. The necking detection method based on geometric feature adaptation according to claim 2, characterized in that, The safety threshold is determined based on the design dimensions of the design graphic.

10. The necking detection method based on geometric feature adaptation according to claim 9, characterized in that, The safety threshold is the design size minus 1-2 minimum length units.

Images