Computer-Generated Multiple Exposure Pattern Division
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current multiple exposure techniques in semiconductor fabrication, such as double patterning, face challenges with complex circuit designs and require extensive rules or long processing times, often necessitating user intervention and using fixed illumination conditions, which are inadequate for resolving densely packed two-dimensional patterns like contact holes.
Innovation Solution
A method involving a computer-generated process to create data for multiple exposure, where a pattern is divided into layers by selecting pattern elements with periodicity, determining effective light sources, and using a pupil filter to optimize illumination conditions for each layer, allowing for improved resolution of densely packed patterns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If rule-based decomposition techniques are used to divide patterns, then pattern division can be performed using pre-established rules, but a large number of rules are required to handle complex circuit design and user intervention is still needed when patterns do not comply with the rules
Solution Approach 1:
The patent replaces rule-based decomposition with model-based decomposition using aerial image computation. Instead of relying on mechanical rule applications, the system uses optical modeling to automatically determine pattern division, eliminating the need for extensive rule sets and user intervention while handling complex circuit designs effectively.
Solution Approach 2:
The patent changes the approach from fixed rule parameters to dynamic illumination condition parameters. By optimizing illumination conditions (such as source shapes and angles) for each divided pattern, the system adapts to complex geometries automatically, removing the need for pre-established rules and user intervention.
2Extent of automation
If model-based decomposition techniques are used to divide patterns, then automated pattern division is achieved, but a long time is required to complete the pattern division process
Solution Approach 1:
The patent segments the pattern division process into multiple exposures with different illumination conditions. By dividing the complex computation into separate aerial image computations for different source shapes and angles, the system achieves automated division while managing computational time through structured segmentation of the analysis process.
Solution Approach 2:
The patent performs preliminary aerial image computations for multiple illumination conditions before final pattern division. By pre-calculating aerial images for different source shapes and angles, the system prepares optimization data in advance, enabling faster and more accurate automated pattern division without excessive processing time during the actual division step.
3Ease of manufacture
If the same illumination condition is used for multiple divided patterns, then the process is simplified, but patterns with dense two-dimensional arrangements cannot be resolved
Solution Approach 1:
The patent applies different illumination conditions to different divided patterns based on their specific geometric characteristics. Instead of using a uniform illumination condition, the system optimizes source shapes and angles locally for each pattern group, enabling resolution of dense two-dimensional arrangements while maintaining process simplicity through systematic differentiation.
Solution Approach 2:
The patent introduces dynamic illumination condition selection for pattern division. By varying source shapes (such as dipole, quadrupole, annular) and angles according to the specific geometry of each divided pattern, the system adapts illumination dynamically to achieve optimal resolution for different pattern types without complicating the overall process.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enables more efficient and automated pattern division, reducing processing time and user intervention, while optimizing illumination conditions for each pattern layer, thereby enhancing the resolution of complex semiconductor patterns.
Implementation Method 1
an exposure apparatus including an illumination optical system which illuminates an original with light from a light source
Implementation Method 2
a projection optical system which projects a pattern of the original onto a substrate
Data Source
AI summary
The present invention provides a method of generating, by a computer, data on patterns of a plurality of originals for use in multiple exposure, in which a single-layer pattern is formed on a substrate by exposing the substrate a plurality of times, in an exposure apparatus including an illumination optical system which illuminates an original with light from a light source, and a projection optical system which projects a pattern of the original onto a substrate.