Wafer Inspection Algorithm Parameter Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current inspection systems for semiconductor wafers require significant user intervention and multiple scans to optimize detection algorithm parameters, leading to inefficiencies and increased time in setting up inspection recipes, especially as the complexity of defect detection increases with smaller device dimensions.
Innovation Solution
A method and system that perform a preliminary scan using default values to detect defects on a wafer without user intervention, select a portion of defects based on a predetermined maximum number, and store information for parameter value determination, allowing for the selection of detection algorithm parameters in a single step without additional scans.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple scans and iterative user intervention are performed to optimize detection algorithm parameters, then parameter selection accuracy is improved, but inspection time and process complexity increase significantly
Solution Approach 1:
The system performs a preliminary scan of the wafer before the actual inspection to collect defect data and automatically determine optimal detection algorithm parameters. This preliminary action eliminates the need for multiple iterative scans and user intervention during the actual inspection process, thereby reducing inspection time while maintaining parameter selection accuracy.
Solution Approach 2:
The system automatically selects detection algorithm parameters using the detected defects from the preliminary scan without requiring user intervention. The parameter selection process serves itself by utilizing the collected defect data to autonomously determine optimal parameters, eliminating time-consuming manual tuning and iterative user involvement.
2Measurement precision
If multiple scans are performed to optimize detection algorithm parameters, then parameter selection accuracy is improved, but process complexity increases
Solution Approach 1:
The optimization process is moved to a preliminary scan phase that occurs before the actual inspection. During this preliminary scan, the system collects defect data and automatically determines optimal parameters, simplifying the actual inspection process to a single scan without iterative complexity.
Solution Approach 2:
The system automatically performs parameter selection using detected defects without requiring user intervention or manual tuning. This self-service approach eliminates complex iterative processes and user-dependent steps, reducing process complexity while maintaining parameter selection accuracy.
3Productivity
If default parameter values are used without optimization, then inspection process speed is improved, but defect detection accuracy deteriorates
Solution Approach 1:
The system performs a preliminary scan to collect defect data and automatically determine optimized parameters before the actual inspection. This allows the use of optimized parameters (improving detection accuracy) while maintaining a single-scan process (preserving inspection speed), resolving the contradiction between speed and accuracy.
4Measurement precision
If user intervention is required for parameter selection, then parameter selection accuracy is improved, but ease of operation deteriorates
Solution Approach 1:
The system automatically selects detection algorithm parameters using defects detected from the preliminary scan, eliminating the need for user intervention. This self-service approach maintains parameter selection accuracy through automated analysis while significantly improving ease of operation by removing manual tuning requirements.
Data Source
AI summary
Methods and systems for generating information to be used for selecting values for parameter(s) of a detection algorithm are provided. One method includes without user intervention performing a scan of an area of a wafer using an inspection system and default values for parameter(s) of a detection algorithm to detect defects on the wafer. The method also includes selecting a portion of the defects from results of the scan based on a predetermined maximum number of total defects to be used for selecting values for the parameter(s) of the detection algorithm. The method further includes storing information, which includes values for the parameter(s) of the detection algorithm determined for the defects in the portion. The information can be used to select the values for the parameter(s) of the detection algorithm to be used for the inspection recipe without performing an additional scan of the wafer subsequent to the scan.