Photomask Defect Detection via Image Averaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for manufacturing semiconductor devices face inefficiencies in defect inspection of photomasks, which can lead to suboptimal performance and increased complexity in semiconductor device production.
Innovation Solution
A method involving the design of a layout, manufacturing a photomask with multiple chip regions, extracting and averaging images from these regions to generate a standard image, and inserting a normal image into the difference regions to detect defects, allowing for efficient comparison and defect detection during the photolithography process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional defect inspection methods are used for photomasks, then the inspection process is simple, but the detection precision and reliability are insufficient
Solution Approach 1:
The photomask inspection process is segmented into multiple steps: extracting first and second images from different chip regions, averaging these images to generate a preliminary standard image, identifying difference regions, and comparing with the original images to detect defects. This segmentation transforms a single complex inspection into manageable sequential operations, improving detection precision while controlling complexity.
Solution Approach 2:
The method creates a standard image by averaging multiple photomask images, effectively generating a reference copy that represents the ideal photomask pattern. This copied standard image is then used for comparison with actual photomask images to detect defects, enhancing measurement precision through reference-based detection.
2Measurement precision
If expensive image modeling tools are used for defect detection, then the defect inspection accuracy improves, but the manufacturing cost increases
Solution Approach 1:
The photomask inspection system performs self-service by automatically generating its own standard image from the photomask images themselves through averaging and processing. This eliminates the need for external expensive image modeling tools, as the system uses its own data to create the reference standard, thereby maintaining high inspection accuracy while reducing manufacturing costs.
Solution Approach 2:
The method changes the parameter of the standard image generation process by using computational averaging of multiple images instead of expensive external modeling tools. This parameter change in the image processing approach maintains defect detection accuracy while significantly reducing the cost投入 required for inspection.
3Reliability
If multiple chip regions are inspected individually, then the defect detection coverage is comprehensive, but the inspection time increases
Solution Approach 1:
The inspection method merges the processing of multiple chip regions by extracting images from different regions (first chip region, second chip region, etc.), averaging them together to create a single preliminary standard image. This merging approach maintains comprehensive defect detection coverage across all regions while reducing inspection time compared to processing each region separately.
Solution Approach 2:
The generated standard image serves as a universal reference for comparing against all chip regions simultaneously. This multi-functional standard image enables comprehensive defect detection across multiple regions without requiring separate inspection processes for each region, thereby maintaining reliability while reducing overall inspection time.
Data Source
AI summary
A method for manufacturing a semiconductor device includes designing a layout, manufacturing a photomask based on the designed layout, and performing a photolithography process using the photomask to form a pattern on a substrate. The manufacturing of the photomask includes preparing the photomask including first and second chip regions, extracting first and second images from the first and second chip regions, respectively, averaging the first and second images to generate a preliminary standard image including a difference region between the first and second images, inserting a normal image into the difference region based on the layout to generate a standard image, and comparing each of the first and second images with the standard image to detect a defect of the first and/or second chip regions.