Multi-mode Inspection System Defect Characterization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current inspection systems for semiconductor manufacturing struggle to effectively detect defects using multiple modes, as they often store only images from modes that detect defects, missing images from modes that do not, and fail to account for co-occurring and correlated attribute/feature values, leading to incomplete defect characterization.
Innovation Solution
A system and method that stores images generated by an inspection system using multiple modes, including those where defects were and were not detected, allowing computer subsystems to identify defects detected in one mode but not another, and determine characteristics based on images from all modes, enabling comprehensive defect characterization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If the inspection system stores only images from modes that detect defects, then storage space is conserved and processing is simplified, but defect characterization becomes incomplete because images from modes that do not detect defects are lost
Solution Approach 1:
The system performs preliminary classification of image modes before storage, identifying which modes are likely to contain defect information. This allows selective storage of relevant images while maintaining complete defect characterization capability, resolving the contradiction between information preservation and storage complexity
Solution Approach 2:
The system applies different storage strategies to different image modes based on their defect detection characteristics. High-value modes that provide unique defect information are stored completely, while redundant modes use compressed or selective storage, optimizing both information retention and storage efficiency
2Measurement precision
If the inspection system uses multiple modes with separate threshold levels, then detection sensitivity for different defect types is improved, but the system cannot account for co-occurring and correlated attribute values across modes
Solution Approach 1:
The system merges threshold determination across multiple image modes by establishing correlated threshold relationships. Instead of managing separate independent thresholds, the system combines threshold logic to evaluate defect candidates consistently across modes, improving detection precision while reducing management complexity
Solution Approach 2:
The system implements feedback mechanisms where detection results from one mode inform threshold adjustments in other modes. This allows the system to account for co-occurring attribute values across modes, as feedback from high-sensitivity modes can optimize thresholds in other modes to maintain overall detection precision
3Measurement precision
If the inspection system performs comprehensive multi-mode analysis for all defect candidates, then defect characterization precision is improved, but processing time and computational resources increase
Solution Approach 1:
The system segments the multi-mode analysis process into priority levels. Critical defect candidates that meet basic criteria in any mode undergo comprehensive multi-mode analysis for precise characterization, while other candidates receive streamlined analysis. This segmentation maintains high precision for important defects while reducing overall processing time
Solution Approach 2:
The system applies partial multi-mode analysis to all defect candidates and reserves comprehensive analysis for high-priority cases. This partial action approach provides sufficient characterization for most defects while avoiding the excessive processing time that would result from applying full multi-mode analysis universally
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach allows for precise defect characterization and differentiation, reducing the need for repeated specimen scans and improving defect detection sensitivity by considering all image attributes across multiple modes, enhancing defect detection and analysis in semiconductor wafer and reticle inspection.
Implementation Method 1
The inspection system is configured for scanning energy over a physical version of the specimen while detecting energy from the specimen to thereby generate the images for the specimen and detect defects on the specimen based on the images
Data Source
AI summary
Methods and systems for determining one or more characteristics for defects detected on a specimen are provided. One system includes one or more computer subsystems configured for identifying a first defect that was detected on a specimen by an inspection system with a first mode but was not detected with one or more other modes. The computer subsystem(s) are also configured for acquiring, from the storage medium, one or more images generated with the one or more other modes at a location on the specimen corresponding to the first defect. In addition, the computer subsystem(s) are configured for determining one or more characteristics of the acquired one or more images and determining one or more characteristics of the first defect based on the one or more characteristics of the acquired one or more images.