SRAF Arrangement Using Coherence Map Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing semiconductor lithography techniques face challenges in ensuring appropriate sub-resolution assist feature (SRAF) arrangement, particularly in rule-based methods where deviations from rules are difficult to manage and model-based methods require extensive time for coherence map calculations, making it hard to derive appropriate SRAF shapes.
Innovation Solution
A method involving a pattern creation apparatus that calculates and combines coherence maps to identify regions of positive and negative coherence, allowing for efficient arrangement of SRAFs at positions with positive coherence, thereby improving lithography margins without extensive calculations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If rule-based SRAF technique is used, then SRAF arrangement follows simple rules, but it is difficult to ensure appropriate lithography margin and hard to manage deviations from rules
Solution Approach 1:
The patent segments the SRAF arrangement process into two distinct phases: first applying rule-based techniques for initial placement, then using model-based optimization only in specific regions where rules fail or deviations occur. This segmentation allows the system to benefit from the simplicity of rule-based methods while applying the precision of model-based methods only where necessary, thus improving lithography margin without sacrificing overall ease of manufacture.
2Manufacturing precision
If model-based SRAF technique is used, then appropriate SRAF shape can be derived, but it takes a lot of time to calculate coherence map
Solution Approach 1:
The patent applies local quality by using model-based coherence map calculations only in specific local regions where rule-based methods produce deviations or insufficient lithography margins, rather than applying model-based methods to the entire design pattern. This localized application significantly reduces calculation time while maintaining high manufacturing precision where it is most needed.
Solution Approach 2:
The patent employs partial action by applying model-based optimization to only a subset of the total SRAF arrangement tasks - specifically those cases where rule-based methods are insufficient. This partial application of the more computationally intensive method achieves the necessary precision without the excessive time cost of applying it universally.
3Adaptability or versatility
If model-based SRAF technique is used, then SRAF can be arranged on difficult patterns, but it is difficult to derive appropriate SRAF shape from coherence map
Solution Approach 1:
The patent introduces an intermediary step that processes the coherence map output to simplify the derivation of appropriate SRAF shapes. This intermediary processing layer translates the complex coherence map data into more manageable information that guides SRAF placement, reducing the complexity of shape derivation while maintaining the ability to handle difficult patterns.
Data Source
AI summary
In an assist pattern arrangement method according to an embodiment, a plurality of patterns, in which an assist pattern is to be arranged, are extracted from a design pattern that is prepared in advance. Then, a resolution map of the extracted pattern is calculated. The resolution map includes a first pattern that increases a resolution of the pattern and a second pattern that decreases the resolution of the pattern. After the resolution map is calculated, a plurality of calculated resolution maps are added. Then, the assist pattern is arranged on the design pattern on the basis of the addition result.