Lithography Lens Model Calibration Data Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for calibrating lens models in lithographic apparatuses are inefficient and suboptimal, particularly due to the large data sets involved and the need for empirical optimization.
Innovation Solution
A method is developed to determine compact lens model dependency data by obtaining input lens model dependency data, determining correctable slit shapes, calculating wavefront shapes, ranking candidate matrices for correction potential and manipulator range usage, and rewriting lens dependencies in a new basis of correctable components.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional lens model calibration methods are used, then comprehensive aberration coverage is achieved, but computational complexity and data processing burden increase significantly
Solution Approach 1:
The patent extracts only the most significant aberration components from the complete set of Zernike polynomials. By identifying and isolating the dominant aberration terms that contribute most to imaging errors, the method reduces the calibration data set size while preserving the essential information needed for accurate lens model calibration, thereby reducing computational complexity without sacrificing measurement precision
Solution Approach 2:
The patent transforms the calibration approach by changing the parameter representation from a complete basis set to a reduced set of dominant components. This parameter change enables the system to work with fewer variables in the lens model calibration process, reducing the computational burden while maintaining the ability to accurately represent the most significant aberrations
2Reliability
If complete Zernike polynomial basis is used for aberration description, then all aberration types are covered, but data set size and processing time increase
Solution Approach 1:
The patent extracts and retains only the dominant aberration components from the complete Zernike polynomial basis. By analyzing which aberration terms contribute most significantly to the overall wavefront error, the method extracts a reduced subset that captures the essential aberration characteristics, thereby reducing processing time while maintaining reliable aberration coverage
Solution Approach 2:
The patent applies partial action by using only the necessary portion of the complete Zernike basis set. Instead of processing all possible aberration terms, the method identifies and processes only the critical subset needed for effective lens calibration, achieving sufficient aberration coverage without the excessive computational burden of complete basis set processing
3Measurement precision
If empirical optimization is performed for lens calibration, then calibration accuracy is improved, but the number of required measurements and computational steps increases
Solution Approach 1:
The patent extracts the most influential aberration components that have the greatest impact on calibration accuracy. By focusing measurements and computations on these dominant components rather than all possible aberration terms, the method maintains high calibration accuracy while significantly improving calibration efficiency by reducing the number of required measurements and computational steps
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Disclosed is a method of determining compact lens model dependency data. The method comprises obtaining input lens model dependency data describing relationships between aberration data and a plurality of manipulators for manipulating elements of a projection system of a lithographic apparatus, said aberration data being described in terms of a first basis; determining correctable slit shapes relating to an exposure slit of the lithographic apparatus from the input lens model dependency data; determining at least one wavefront shape for each said correctable slit shape to determine a candidate matrix; determining a plurality of correctable components by ranking said candidate matrix in terms of correction potential and/or manipulator range usage, each said correctable component defining a basis function of a second basis, said second basis being different to said first basis; and determining compact lens model dependency data in terms of said second basis.