EUV Mask Inspection With Gray Level Compensation for False Positives
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The semiconductor industry faces challenges in accurately inspecting EUV masks due to false positives from brightness changes caused by EUV exposure, leading to reduced tool availability and increased downtime.
Innovation Solution
A system and method for inspecting EUV masks that involves collecting a brightness map and compensating for local brightness changes using a transform equation and machine learning models, along with material type identification from GDS files to reduce false positives.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If EUV masks are exposed repeatedly to EUV light during photolithography operations, then production efficiency increases, but brightness changes occur on the mask surface causing false positive defect detections
Solution Approach 1:
The system performs preliminary actions by collecting a brightness map of the mask before inspection and generating a compensated image using transform functions that account for expected brightness changes. This preliminary compensation prepares the inspection data to resist the harmful brightness changes that occur during repeated EUV exposure, thereby maintaining inspection accuracy while enabling continuous production.
Solution Approach 2:
The system implements feedback by comparing the compensated inspection image with a reference image and using the results to adjust and refine the brightness compensation model. The feedback loop continuously improves the accuracy of brightness change prediction and compensation, ensuring that false positives are minimized even as the mask undergoes repeated exposure cycles.
2Device complexity
If traditional inspection methods are used without brightness compensation, then device complexity remains low, but false positive defect detections increase leading to reduced tool availability
Solution Approach 1:
The system introduces an intermediary brightness compensation layer between the raw inspection image and the defect detection process. A transform function acts as a mediator that mathematically adjusts the inspection image to account for brightness changes, separating the harmful brightness variation from the actual defect signal. This intermediary step reduces false positives without requiring complete redesign of the inspection system.
Solution Approach 2:
The system applies parameter changes by using a transform function to modify the brightness parameters of the inspection image. The transform adjusts key parameters such as gray level values and intensity distribution to compensate for exposure-induced brightness changes, enabling accurate defect detection while maintaining relatively simple system architecture.
3Reliability
If gray level map local compensation is applied to compensate for brightness changes, then false positive defect detections are reduced, but processing time and computational resources increase
Solution Approach 1:
The system applies segmentation by dividing the mask inspection process into distinct regions and applying localized compensation to different areas. Instead of processing the entire mask uniformly, the system segments the brightness compensation based on local characteristics, which reduces the computational burden while maintaining high accuracy in detecting actual defects versus brightness artifacts.
Solution Approach 2:
The system employs partial action by applying brightness compensation selectively to regions where brightness changes are most problematic, rather than uniformly across the entire mask. This approach achieves sufficient defect detection accuracy without the full computational overhead of complete mask compensation, thereby reducing processing time while maintaining reliability.
Data Source
AI summary
A method includes: generating a first image by scanning a mask; following the generating a first image, exposing the mask in a photolithography operation; following the exposing, generating a second image by scanning the mask; generating a compensated third image by performing a gray level map local compensation on the second image, the gray level map local compensation being via a transform function; generating a comparison result by comparing the compensated third image with the first image; and based on the comparison result, determining whether the mask has a defect thereon.