Wafer Defect Detection Using Localized Reference Images
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current defect detection methods for wafers face challenges due to wafer-to-wafer noise and die-to-die noise, which decrease accuracy, especially when standard die images are used, as they may not adequately represent noise sources on the wafer.
Innovation Solution
A method that determines the locations of weak geometries in a wafer design, scans the wafer with a detection system, and detects defects based on output from multiple instances of these geometries within a single die, using computer systems to align and compare images for improved accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If standard die images are used for defect detection, then the inspection process can be simplified, but wafer-to-wafer noise and die-to-die noise increase, decreasing defect detection accuracy
Solution Approach 1:
The patent segments the wafer surface into multiple regions of interest (ROIs) and processes each region separately using localized reference images. This segmentation allows noise characteristics to be captured locally for each region, reducing the impact of wafer-to-wafer and die-to-die noise while maintaining a manageable inspection process.
Solution Approach 2:
The patent applies local quality by using region-specific reference images instead of a single global standard die image. Each ROI has its own reference image that captures the local noise characteristics, allowing for more accurate defect detection in each specific region while keeping the overall process efficient.
2Productivity
If standard die images from other wafers are used, then inspection can proceed without additional overhead, but wafer-to-wafer noise interferes with defect detection
Solution Approach 1:
The patent performs preliminary action by capturing reference images from the actual wafer being inspected during setup, rather than relying on pre-stored images from other wafers. This preliminary capture of wafer-specific characteristics eliminates wafer-to-wafer noise while maintaining efficient inspection throughput.
3Ease of manufacture
If design data is used to generate standard die images, then the process is efficient, but the images may not adequately represent actual noise sources on the wafer
Solution Approach 1:
The patent uses copying by capturing actual optical images of the wafer surface during setup and using these real-world copies as reference images, rather than relying on theoretical design data. This ensures the reference images accurately represent actual noise sources and manufacturing variations while keeping the process efficient.
Data Source
AI summary
Systems and methods for detecting defects on a wafer are provided. One method includes determining locations of all instances of a weak geometry in a design for a wafer. The locations include random, aperiodic locations. The weak geometry includes one or more features that are more prone to defects than other features in the design. The method also includes scanning the wafer with a wafer inspection system to thereby generate output for the wafer with one or more detectors of the wafer inspection system. In addition, the method includes detecting defects in at least one instance of the weak geometry based on the output generated at two or more instances of the weak geometry in a single die on the wafer.