Outline Image Alignment for Ultrafine Pattern Inspection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing pattern inspection apparatuses face challenges in achieving high-precision alignment between outlines of ultrafine patterns on semiconductor wafers, leading to prolonged processing times and reduced yield in LSI manufacturing due to complex alignment processing.
Innovation Solution
A pattern inspection apparatus and method that generate actual and reference outline images using a predetermined function to define gray scale values based on distance from pixel centers to outline positions, enabling precise alignment through gray scale differences between these images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional alignment processing between images is used (minimizing luminance deviation by least squares method), then alignment can be performed relatively quickly, but alignment precision between outlines is insufficient for ultrafine patterns
Solution Approach 1:
The patent replaces conventional image-based alignment (using luminance values and least squares method) with outline-based alignment (using edge detection and differential values). This substitution changes the fundamental approach from pixel intensity comparison to geometric feature matching, achieving higher precision for ultrafine patterns while maintaining reasonable processing time through efficient differential calculation methods
Solution Approach 2:
The patent changes the alignment parameters from luminance values to differential values (first and second differentials of concentration values). By transforming the alignment criterion from intensity-based to gradient-based parameters, the method achieves superior outline alignment precision necessary for 10nm-scale patterns, resolving the contradiction between precision and processing time
2Measurement precision
If high-precision outline alignment is performed using differential methods, then alignment accuracy improves, but processing time increases significantly
Solution Approach 1:
The patent performs preliminary edge candidate extraction using Sobel filters before the main alignment processing. By pre-identifying edge candidates and preparing differential value calculations in advance, the method reduces the computational burden during the actual alignment process, thereby improving inspection throughput while maintaining high outline alignment accuracy
Solution Approach 2:
The patent segments the alignment process into distinct stages: edge candidate identification using Sobel filters, differential value calculation for identified edges, and final alignment computation. This segmentation allows optimized processing at each stage, preventing the computational complexity from bottlenecking overall productivity while ensuring accurate outline alignment
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 highly accurate alignment with reduced processing time, improving the efficiency and accuracy of defect detection in ultrafine pattern inspection.
Implementation Method 1
there has been developed another inspection apparatus that acquires a pattern image by scanning an inspection target substrate with primary electron beams and detecting secondary electrons emitted from the inspection target substrate due to the irradiation with the primary electron beams
Data Source
AI summary
A pattern inspection apparatus includes an actual outline image generation circuit to generate an actual outline image of a predetermined region defined by a function, where the gray scale value of each pixel in the predetermined region including plural actual image outline positions on an actual image outline of a figure pattern in an inspection image is dependent on a distance from the center of a pixel concerned to the closest actual image outline position in the plural actual image outline positions, and a reference outline image generation circuit to generate a reference outline image of the predetermined region defined by the function, where a gray scale value of each pixel in the predetermined region is dependent on a distance from the center of a pixel concerned to the closest reference outline position in plural reference outline positions on a reference outline to be compared with the actual image outline.