Golden Grid Reference Image for Semiconductor Inspection Alignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current inspection methods for semiconductor manufacturing face challenges in accurately aligning test images to design due to pixel location variations from scan to scan, leading to unstable defect detection and attribute stability.
Innovation Solution
A system and method that acquire a reference image, modify it to fit a design grid to generate a golden grid image, and use this image for aligning test images, thereby stabilizing defect detection and reducing nuisance filtering variability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If pixel to design alignment (PDA) offsets are used to align test images to design, then alignment can be performed, but alignment accuracy varies from scan to scan, mode to mode, and tool to tool due to process and specimen variation
Solution Approach 1:
The patent creates a copied reference image from a previous scan that has already been aligned to the design grid. This reference image serves as a stable template for aligning subsequent test images, eliminating the need for direct PDA alignment with the original design for each scan. The reference image copy preserves the correct geometric relationships while providing a consistent alignment target across multiple scans and tools.
Solution Approach 2:
The patent performs alignment to the design grid in advance during the creation of the reference image, rather than performing PDA alignment for each test image during runtime. This preliminary alignment action creates a stable reference that can be reused, avoiding repeated alignment operations that are subject to variation.
2Productivity
If RTCM care areas are generated from design and applied to test images using PDA offsets, then defect detection can be performed, but RTCM region map changes from scan to scan causing defect count variation
Solution Approach 1:
The patent copies the RTCM care areas from the reference image (which is already aligned to design) and applies them to test images aligned to the reference image. This creates a consistent care area mapping that does not change from scan to scan, eliminating the source of defect count variation while maintaining defect detection capability.
3Difficulty of detecting and measuring
If reference patch-based attributes are computed from defects, then defect characterization can be performed, but attributes are not stable due to scan to scan variation
Solution Approach 1:
The patent copies reference patch images from the reference image (which is stable and aligned to design) rather than computing them from each test image. This ensures that defect attributes such as reference context pattern and reference contrast are derived from consistent reference data, making them stable across scans while maintaining characterization capability.
Data Source
AI summary
Methods and systems for setting up inspection of a specimen are provided. One system includes one or more computer subsystems configured for acquiring a reference image for a specimen and modifying the reference image to fit the reference image to a design grid thereby generating a golden grid image. The one or more computer subsystems are also configured for storing the golden grid image for use in inspection of the specimen. The inspection includes aligning a test image of the specimen generated from output of an inspection subsystem to the golden grid image.