Sub-Resolution Assist Feature Inspection Thresholds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional reticle inspection methods are inadequate for detecting real defects in sub-resolution assist features (SRAFs) on semiconductor wafers, often identifying nuisance defects and missing critical ones.
Innovation Solution
A method that uses different sensitivity levels and threshold values for SRAF boundaries compared to other reticle features, allowing for separate flux measurement and inspection of SRAFs with adjustable sensitivity, and employing reference flux measurements from a design database to determine if SRAFs are oversized or undersized.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional inspection methods with single threshold are used, then the inspection process is simple, but it detects nuisance defects and fails to capture real defects in SRAFs
Solution Approach 1:
The patent applies local quality by implementing different inspection thresholds and sensitivity levels for different regions and feature types. Specifically, it uses separate threshold values for SRAF boundaries versus other reticle features, and different threshold criteria for oversized versus undersized SRAFs. This localized differentiation allows the inspection system to optimize detection accuracy for each specific inspection target without applying a one-size-fits-all approach, thereby reducing nuisance defects while capturing real defects.
Solution Approach 2:
The patent segments the inspection process into multiple independent evaluation stages: (1) identifying SRAF boundary areas versus other features, (2) measuring flux for each region, (3) comparing against different threshold sets for oversized/undersized conditions, and (4) making defect determinations. This segmentation allows complex defect detection to be broken down into manageable steps with specialized criteria for each stage, improving overall reliability while maintaining systematic simplicity.
2Reliability
If high sensitivity inspection is applied to all features, then real defects are detected, but nuisance defects are also identified
Solution Approach 1:
The patent implements local quality by applying different sensitivity levels and threshold criteria to different feature types. High sensitivity inspection is applied specifically to SRAF boundary areas with appropriate threshold values, while other reticle features are inspected with different criteria. This localized high sensitivity approach ensures real defects in SRAFs are detected without triggering nuisance defect alarms in other regions, as each area is evaluated against its appropriate threshold set.
Solution Approach 2:
Instead of applying uniform high sensitivity across the entire reticle and then filtering results, the patent inverts the approach by first identifying the specific regions of interest (SRAF boundaries) and applying high sensitivity only there. The inspection system treats SRAF areas differently from other features, using specialized threshold values that are less likely to trigger nuisance defects while maintaining high detection capability for real SRAF defects.
Data Source
AI summary
Disclosed are methods and apparatus for inspecting a sub-resolution assist features (SRAF) on a reticle. A test flux measurement for a boundary area that encompasses a width and a length portion of a test SRAF is determined, and at least one reference flux measurement for one or more boundary areas of one or more reference SRAF's is determined. The test flux measurement is compared with the reference flux measurements. The comparison is used to then determine whether the test SRAF is undersized or oversized. If the test SRAF is determined to be oversized, it may then be determined whether the test SRAF is defective based on the comparison using a first threshold.