Defect Pattern Evaluation via Occurrence Rate Calculation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for evaluating pattern defects in semiconductor manufacturing struggle to accurately extract and prioritize defects with low detection numbers but high importance, leading to potential missed critical defects and reduced manufacturing yield.
Innovation Solution
A defect pattern evaluation method and apparatus that calculates and ranks defect occurrence rates by pattern layout, using a combination of optical inspection, scanning electron microscopy, and design-based binning to identify and prioritize systematic defects based on their probability of occurrence rather than mere detection numbers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If patterns are classified and ranked based on the number of defects detected, then the quantity of defects is improved, but the accuracy of extracting important defects with low detection numbers deteriorates
Solution Approach 1:
The patent changes the evaluation parameter from defect detection number to defect occurrence rate. By calculating the ratio of defect detection number to pattern arrangement number, the system transforms the assessment metric to account for both the frequency of defects and the total number of patterns, thereby accurately identifying important defects regardless of their absolute detection count
Solution Approach 2:
The patent segments the evaluation process into two distinct components: defect detection (quantifying defects) and pattern arrangement counting (quantifying total patterns). By separately measuring these elements and combining them through division, the system achieves a more nuanced evaluation that prevents important but low-count defects from being overlooked
2Productivity
If focus on high detection number patterns, then productivity of defect review is improved, but reliability of detecting critical systematic defects deteriorates
Solution Approach 1:
The patent changes the prioritization parameter from defect detection number to defect occurrence rate. This parameter transformation enables the system to identify critical systematic defects based on their relative frequency within pattern types, ensuring that even defects with low absolute counts but high occurrence rates are prioritized for review
Solution Approach 2:
The patent implements a feedback mechanism where defect occurrence rates are calculated and used to adjust the review priority. By continuously evaluating the ratio of defects to total patterns and using this information to guide subsequent review efforts, the system ensures that resources are allocated to the most critical defect types while maintaining high productivity
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 the accurate extraction and evaluation of important pattern defects with low detection numbers, enhancing process margin evaluation and improving manufacturing yield by focusing on high-risk defects, even if they occur infrequently.
Implementation Method 1
defects are detected by performing optical defect inspection on a pattern on a substrate
Implementation Method 2
using a combination of optical inspection, scanning electron microscopy
Data Source
AI summary
According to a defect pattern evaluation method of an embodiment, defects are detected by performing optical defect inspection on a pattern on a substrate. Then, the defects are classified according to a type of a pattern layout using a pattern layout corresponding to coordinates of the defects. Further, a computer calculates a defect occurrence rate by dividing the number of defects of each pattern layout by an arrangement number of the pattern layouts in an inspection region. Then, the defect occurrence rate of each pattern layout is output as an evaluation result.