Mark Layout Optimization for Lithography Throughput
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current lithographic apparatuses face challenges in accurately placing patterns on substrates due to imperfections and the need for extensive measurement operations, which limit throughput and increase costs, particularly in achieving precise overlay and focus uniformity as feature sizes decrease.
Innovation Solution
A method for determining an optimal layout of mark positions on a patterning device or substrate using a model-based approach that reduces the number of mark positions while maintaining model uncertainty within acceptable limits, employing a combination of initial mark layout reduction and model uncertainty metric analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If extensive measurement operations are performed to achieve accurate overlay and focus uniformity, then measurement precision is improved, but productivity deteriorates due to limited throughput
Solution Approach 1:
The patent extracts only the essential mark positions needed for accurate process fingerprint determination from the complete set of available marks. By identifying and removing redundant marks through model uncertainty analysis, the method retains sufficient measurement precision while significantly reducing the number of marks that need to be measured, thereby improving throughput.
Solution Approach 2:
Instead of measuring all available marks (excessive action), the method performs partial measurement by selecting only the critical subset of marks that provide sufficient model accuracy. The model uncertainty metric determines when partial measurement is adequate, allowing the process to achieve acceptable precision without the overhead of complete measurement.
2Productivity
If the number of mark positions is reduced to improve throughput, then productivity is improved, but measurement precision deteriorates
Solution Approach 1:
The patent implements feedback through model uncertainty metrics that continuously evaluate the quality of the process fingerprint model based on the selected mark positions. This feedback mechanism allows the system to determine whether reducing the number of marks has compromised measurement precision, enabling dynamic adjustment to maintain adequate accuracy while maximizing throughput.
Solution Approach 2:
The method changes the parameter of mark position selection from a fixed comprehensive set to a dynamically optimized subset. By using model uncertainty as a criterion, the system adapts the number and positions of marks based on the specific process conditions and requirements, achieving optimal balance between precision and productivity.
3Manufacturing precision
If advanced alignment models and focus models are used to model distortions, then manufacturing precision is improved, but device complexity increases due to increased number of mark positions required
Solution Approach 1:
The patent extracts only the critical mark positions necessary to constrain the advanced alignment and focus models effectively. By identifying which marks contribute most to model accuracy and removing redundant ones, the method maintains manufacturing precision while reducing the complexity of the mark layout and measurement system.
Data Source
AI summary
A method for determining a layout of mark positions across a patterning device or substrate, the method including: obtaining a model configured to model data associated with measurements performed on the patterning device or substrate at one or more mark positions; obtaining an initial mark layout including initial mark positions; reducing the initial mark layout by removal of one or more mark positions to obtain a plurality of reduced mark layouts, each reduced mark layout obtained by removal of a different mark position from the initial mark layout; determining a model uncertainty metric associated with usage of the model for each reduced mark layout out of the plurality of reduced mark layouts; and selecting one or more reduced mark layouts based on its associated model uncertainty metric.