Wafer Defect Signature Detection via Density Map Clustering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional in-process monitoring techniques for wafer inspection face inefficiencies in defect detection, particularly in the second phase where only a small fraction of defects are reviewed in detail, leading to resource wastage and incomplete analysis of defects caused by single physical occurrences.
Innovation Solution
A system and method for signature detection that includes a processor with modules for acquiring defect density maps, determining thresholds, clustering sections into signatures, and unifying signatures across different resolution levels, enabling the identification and characterization of defect signatures efficiently.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If only a small fraction of defects are reviewed in the second phase, then resource consumption is reduced, but defect analysis completeness deteriorates
Solution Approach 1:
The patent segments defects into different groups based on their characteristics and defect density patterns. By dividing the large set of defects into meaningful segments (signatures), the system can selectively review representative defects from each segment rather than reviewing all defects, thus reducing resource consumption while maintaining analysis completeness.
Solution Approach 2:
The patent creates defect density maps that serve as simplified representations or copies of the actual defect distribution. These maps capture the essential patterns and characteristics of defects without requiring detailed analysis of every individual defect, allowing resources to be focused on analyzing representative samples while the maps provide comprehensive coverage.
2Reliability
If a large number of defects are detected in the first phase, then defect detection coverage is improved, but processing efficiency deteriorates
Solution Approach 1:
The patent performs preliminary actions by generating defect density maps and identifying defect signatures in the first phase. This preliminary analysis organizes and pre-processes the defect data, so that in the second phase, only representative defects from identified signatures need detailed review, significantly improving processing efficiency while maintaining comprehensive detection coverage.
Solution Approach 2:
The patent extracts key defect signatures and characteristic patterns from the large set of detected defects. By taking out only the essential signature information and representative defects for detailed review, the system maintains comprehensive detection coverage while reducing the processing burden in the second phase.
3Measurement precision
If defects from a single physical occurrence are analyzed individually, then defect detection accuracy is improved, but resource allocation deteriorates
Solution Approach 1:
The patent merges multiple defects that originate from the same physical occurrence into a single defect signature. By combining these related defects and analyzing them as one unified entity, the system maintains accurate detection of the underlying physical issue while avoiding redundant analysis of each individual defect manifestation, thus optimizing resource allocation.
Data Source
AI summary
Methods, systems, and computer program products for signature detection. One example of a method includes: acquiring an article defect density map comprising a plurality of sections corresponding to a first resolution level which is indicative of defect numbers for the sections, and determining a distribution representative of the defect numbers or function thereof; determining a threshold in accordance with said distribution, and identifying sections, out of said plurality of sections in the article defect density map, with defect numbers or function thereof above the threshold; and clustering at least part of adjoining identified sections, into one or more signatures, thus detecting said one or more signatures.