EUV Reticle Repeater Defect Detection via Stacked Difference Images
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current repeater defect detection methods for reticles, particularly in extreme ultraviolet (EUV) lithography, struggle to accurately detect smaller defects due to noise issues arising from differences in setup and test wafers, leading to missed defects and false positives.
Innovation Solution
A system and method that generates stacked difference images from multiple instances of patterned areas on a wafer, using a computer subsystem to detect defect candidates only when they are consistently present across images, thereby reducing noise and improving sensitivity for EUV reticle defect detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a setup wafer is used for image alignment in repeater defect detection, then image alignment can be performed, but noise is introduced due to process condition variations between setup and test wafers
Solution Approach 1:
The patent uses multiple test wafers printed with the same reticle to create reference images, eliminating the need for a separate setup wafer. By copying the actual test conditions across multiple wafers and using their average as the reference, the system avoids introducing noise from setup wafer variations while maintaining accurate image alignment
Solution Approach 2:
The patent combines images from multiple test wafers to create a composite reference image. By merging data from multiple wafers that have all undergone the same process conditions, the system achieves better alignment accuracy and reduces noise compared to using a single setup wafer
2Measurement precision
If noise reduction methods are applied to detect repeater defects, then fewer nuisances are detected, but smaller EUV repeater defects cannot be detected
Solution Approach 1:
The patent extends defect detection from a single-wafer two-dimensional analysis to a multi-wafer three-dimensional analysis. By stacking difference images from multiple wafers, the system adds a temporal/dimensional dimension that enhances noise reduction while preserving small defect signals that appear consistently across all wafers
Solution Approach 2:
The patent replaces traditional mechanical noise filtering methods with a statistical approach using multiple wafer measurements. Instead of applying fixed thresholds or filters that may eliminate small defects, the system uses the variability across multiple wafers to distinguish noise from real defects, preserving sensitivity to smaller EUV defects
3Productivity
If repeater defect detection is performed as a post-processing operation, then normal die-to-die defect detection can be completed first, but the detection sensitivity for repeater defects is reduced
Solution Approach 1:
The patent performs repeater defect detection continuously during the inspection process rather than as a separate post-processing step. By integrating the repeater detection algorithm into the main inspection workflow and performing it in real-time as images are acquired, the system maintains high sensitivity while improving overall productivity
Data Source
AI summary
Systems and methods for detecting defects on a reticle are provided. One system is configured for generating different stacked difference images for multiple instances of first patterned areas in different rows on a wafer based on images generated for the first patterned areas in the different rows. The system is also configured for performing double detection based on the different stacked difference images. The system then identifies defects on the reticle based on the defects detected by the double detection. As described further herein, the systems and methods detect defects from multiple reticle rows printed on a wafer, which can reduce noise and enable detection of substantially small repeater defects. The embodiments are particularly useful for high sensitivity repeater defect detection for extreme ultraviolet (EUV) reticles and multi-die reticles (MDR).