Photomask OPC Hashing for Consistent Pattern Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current semiconductor device manufacturing methods face challenges in achieving consistent optical proximity correction, leading to distorted patterns and reduced integration and reliability due to interference and diffraction effects during photolithography, which affect the accuracy and consistency of pattern formation on substrates.
Innovation Solution
The method involves performing optical proximity correction (OPC) by dividing design patterns into segments, generating hash values for each segment, and applying a correction bias to segments with the same hash value, ensuring consistent correction patterns are applied to segments with similar surrounding environments, thereby reducing pattern distortion and improving accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If optical proximity correction is performed without segmentation and hashing, then the correction process can be simpler, but the consistency and accuracy of correction across similar segments deteriorates
Solution Approach 1:
The design pattern is divided into multiple segments along the contour, allowing independent correction analysis for each segment. This segmentation enables the system to identify and apply consistent correction biases to similar segments while maintaining the ability to handle complex pattern variations, thereby improving manufacturing precision without overwhelming complexity
Solution Approach 2:
The invention introduces hash values as a parameter to classify segments with similar surrounding environments. By transforming the complex spatial relationship information into discrete hash values, the system can efficiently identify and apply consistent correction biases, improving pattern formation accuracy while managing computational complexity through parameterization
2Manufacturing precision
If correction bias is applied uniformly to all segments, then the process is simpler, but the accuracy for segments with different surrounding environments deteriorates
Solution Approach 1:
The invention applies different correction biases to different segments based on their local surrounding environments, as identified by hash values. This local quality approach ensures that each segment receives the appropriate correction for its specific context, improving correction accuracy while maintaining systematic management through the hashing mechanism
Solution Approach 2:
The invention creates a hash value representation of the surrounding environment for each segment, effectively creating a simplified copy or fingerprint of the local context. This allows the system to identify identical environments and apply consistent correction biases, improving accuracy for segments with similar environments while managing complexity through the copying approach
3Manufacturing precision
If detailed individual analysis is performed for each segment, then correction accuracy improves, but execution time increases
Solution Approach 1:
The invention performs preliminary hashing of segment surrounding environments before the actual correction analysis. By pre-classifying segments into groups with identical hash values, the system reduces the computational burden during the correction phase, maintaining high correction consistency while significantly reducing execution time through this preliminary categorization action
Solution Approach 2:
The invention merges segments with identical surrounding environments into groups identified by the same hash value. This merging allows the system to perform correction analysis once per unique environment type and apply the result to all segments in that group, maintaining correction consistency while reducing overall execution time through combined processing
Data Source
AI summary
A photomask fabrication method comprising performing an optical proximity correction (OPC) on a design pattern to generate a correction pattern, and manufacturing the photomask having the correction pattern. The step of performing the OPC includes dividing the design pattern into a plurality of segments, producing a hash value for each of the plurality of segments, and generating the correction pattern by applying a first correction bias to ones of the plurality of segments having a same hash value, wherein at least two of the plurality of segments have the same hash value. The step of producing the hash value includes generating a key segment in a target segment, creating a query region around the key segment, and producing the hash value for the target segment based on a pattern image in the query region.