Nuisance Reduction Using Location-Based Attributes in Semiconductor Images
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In semiconductor manufacturing, the shrinking dimensions of critical features and increasing wafer size pose challenges in defect detection, particularly with metal lines, where previous techniques are inaccurate due to resolution limits and time-consuming, leading to high nuisance rates that can misidentify defects or obscure critical issues.
Innovation Solution
A method using x-axis and y-axis pixel grey level intensity charts to classify potential defects relative to material lines, determining their position and distance, and distinguishing between defects of interest and nuisance events with improved precision and speed, enabling accurate identification and localization of defects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If pattern-based attributes are used to classify defects, then defect classification can be performed, but measurement precision deteriorates due to resolution limits causing inaccurate defect localization
Solution Approach 1:
The patent segments the defect classification process into two distinct stages: first, pattern-based attributes are used for initial defect identification and filtering; second, location-based attributes are applied to precisely determine defect positions relative to metal lines. This segmentation allows each method to be optimized for its specific purpose, resolving the contradiction between classification capability and localization accuracy.
Solution Approach 2:
The patent introduces location-based attributes as an intermediary mechanism that bridges the gap between pattern-based classification and precise defect localization. By calculating distances from defect centroids to metal line edges using projection data, the system achieves accurate defect positioning without being limited by optical resolution constraints.
2Reliability
If very small care areas are defined around nuisance structures to filter nuisance events, then nuisance reduction can be achieved, but device complexity increases due to time-consuming rule creation
Solution Approach 1:
The patent implements a self-service approach where the system automatically determines care area boundaries by calculating distances from metal line edges based on projection data, eliminating the need for manual rule creation. The care areas are dynamically generated for each image based on the actual metal line positions, reducing both complexity and processing time while maintaining effective nuisance filtering.
Solution Approach 2:
The patent changes the parameter used to define care areas from fixed predetermined values to dynamic values based on actual metal line positions and projection data. By using location-based attributes that adapt to each specific image configuration, the system achieves effective nuisance reduction without requiring complex manual rule creation.
3Productivity
If optical imaging is used to detect defects on metal lines, then defect detection can be performed, but measurement precision deteriorates due to resolution limits causing blurry line appearance
Solution Approach 1:
The patent introduces projection data as an intermediary that captures precise metal line position information without being affected by optical resolution limits. By projecting image data along scan directions and calculating distances from defect centroids to projection peaks, the system achieves accurate line position determination even when the original optical images are blurry.
Solution Approach 2:
The patent replaces direct optical measurement of metal line positions with a computational approach using projection data and distance calculations. This substitution of mechanical/optical measurement with computational analysis eliminates the resolution limits of optical systems while maintaining defect detection capability.
Data Source
AI summary
Methods and systems are disclosed that provide nuisance reduction in images, such as semiconductor images that include one or more metal lines. A potential defect is correlated against pixel grey level intensity charts for two perpendicular axes. A position of the potential defect relative to a pattern, such as a metal line, is determined along the two axes. The potential defect can be classified as a defect of interest or nuisance event.