Metrology Reference Image Orthogonalization for Crosstalk Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Metrology techniques in lithographic processes are vulnerable to crosstalk from neighboring structures, which affects the accuracy of measurements such as overlay, focus, and dose determination.
Innovation Solution
A method is introduced to determine an orthonormalized structure of interest reference image to correct for the impact of nuisance structures by orthonormalizing the structure of interest reference image with respect to nuisance structure reference images, and using this correction to improve measurement accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If metrology measurements are performed on structures in lithographic processes, then process control and verification are achieved, but measurement accuracy deteriorates due to crosstalk from neighboring structures
Solution Approach 1:
The patent segments the measurement process into multiple components: obtaining a measured image of the structure of interest, obtaining a nuisance structure image representing crosstalk, and then processing these separately before combining them. This segmentation allows the crosstalk component to be identified and removed, improving measurement accuracy while maintaining process control capability
Solution Approach 2:
The patent extracts the nuisance structure image (crosstalk) from the overall measurement system and treats it as a separate component. By obtaining a dedicated nuisance structure image and using it to correct the measured image, the harmful crosstalk effect is isolated and removed, thereby improving measurement precision without sacrificing the ability to perform process control
2Measurement precision
If crosstalk correction is applied to metrology images, then measurement accuracy is improved, but processing complexity increases
Solution Approach 1:
The patent performs preliminary action by obtaining the nuisance structure image before correcting the measured image. The nuisance structure image is obtained in advance and can be reused for multiple measurements, which reduces the overall processing complexity compared to performing complex corrections for each individual measurement. This preliminary preparation simplifies the subsequent correction process
Solution Approach 2:
The patent creates a copy of the nuisance structure effect through the nuisance structure image, which represents the crosstalk that would affect the measured image. By working with this copied representation rather than directly manipulating the complex physical crosstalk interactions, the processing becomes more manageable and systematic, improving accuracy without excessive complexity
Data Source
AI summary
Disclosed is a method of determining an orthonormalized structure of interest reference image. the orthonormalized structure of interest reference image for applying to a measured image of the structure of interest to correct for the effect of crosstalk from at least one nuisance structure. The method comprises determining a structure of interest reference image based on knowledge of the structure of interest; determining at least one nuisance structure reference image based on knowledge of the at least one nuisance structure; and orthonormalizing the structure of interest reference image to the at least one nuisance reference image to obtain the orthonormalized structure of interest reference image.