Automated Reticle Defect Detection in Semiconductor Lithography
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional lithography processes face challenges in detecting and addressing repeating defects in reticle masks, which lead to high yield losses due to the manual and time-consuming nature of defect identification and the inability to promptly stop the manufacturing process when defects are detected.
Innovation Solution
An automated image capturing and processing technique that transforms pixel domain images of semiconductor wafers into a transform domain to identify repeating pattern defects, allowing for real-time detection and immediate cessation of the stepper process with a defective reticle, followed by its replacement with a defect-free one.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If manual defect identification methods are used in lithography processes, then detection simplicity is maintained, but detection time increases and yield losses occur due to delayed defect detection
Solution Approach 1:
The patent replaces manual visual inspection with an automated optical imaging system that captures wafer images, processes them through defect detection algorithms, and automatically identifies repeating defects. This substitution of mechanical/manual processes with automated optical-electronic systems directly reduces detection time while implementing the required automation.
Solution Approach 2:
The system enables self-detection by automatically capturing images, processing defect information, and generating defect maps without requiring continuous manual intervention. The automated pipeline allows the system to service itself in terms of defect identification, reducing both time loss and the need for manual labor.
2Productivity
If repeating defects in reticle masks are not detected, then production continues uninterrupted, but yield losses increase due to defective devices being manufactured
Solution Approach 1:
The system establishes a feedback loop where defect detection results are fed back to the manufacturing process. When repeating defects are detected, the system provides feedback that triggers alerts and prevents further production with the defective reticle, thus maintaining productivity while protecting yield rate through continuous monitoring and response.
Solution Approach 2:
The system performs preliminary defect detection before defective devices are mass-produced. By detecting repeating defects early in the process and implementing preventive measures, the system takes preliminary action to avoid the harmful consequences of manufacturing defects, thereby protecting both productivity and yield rate.
3Device complexity
If conventional lithography processes are used, then manufacturing simplicity is maintained, but the ability to detect and respond to repeating defects is insufficient
Solution Approach 1:
The defect detection system is segmented into distinct functional modules: image capture, defect information processing, defect map generation, and analysis. This segmentation allows each module to be optimized independently for its specific function, improving overall detection precision while managing system complexity through modular design.
Solution Approach 2:
The system transitions from conventional two-dimensional visual inspection to three-dimensional defect analysis by capturing images at multiple focal planes and processing them to generate comprehensive defect maps. This dimensional enhancement improves detection precision by revealing defects that may be invisible in single-plane images, while the automated processing manages the increased complexity.
Data Source
AI summary
A method includes capturing an image of the pattern using one or more scans across a surface of the partially completed wafer. The method includes processing information associated with the captured image of the pattern in a first format (e.g., pixel domain) into a second format, e.g., transform domain. The method includes determining defect information associated with the image of the pattern in the second format and processing the defect information (e.g., wafer identification, product identification, layer information, x-y die scanned) to identify at least one defect associated with a spatial location of a repeating pattern on the partially completed wafer provided by a reticle. The method includes identifying the reticle associated with the defect and a stepper associated with the reticle having the defect and ceasing operation of the stepper. The damaged reticle is replaced, and the process resumes using a replaced reticle.