Pattern Matching with Integrated Width Checks in Multi-Patterned Memory Cells
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing pattern matching methods struggle to accurately match patterns formed by multi-patterning, particularly in areas prone to pattern dimension variations, such as memory cells, due to issues like pitch shifts and deformation, making it difficult to optimize and monitor process fluctuations.
Innovation Solution
A pattern matching method that involves generating images of specific regions, performing initial matching on edge regions, classifying patterns into groups based on layer information, measuring and integrating pattern widths, and comparing integrated values to ensure correct matching across different regions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If pattern matching is performed in the central part of the memory cell to avoid edge deformation, then pattern matching accuracy is improved, but it becomes difficult to judge whether the matching is correct due to pitch shifts
Solution Approach 1:
The patent divides the pattern matching process into multiple stages: first matching reference patterns at the edge region to establish a baseline, then matching patterns in the central region. This segmentation allows the system to use edge region information to verify central region matching results, resolving the contradiction between matching accuracy and verification difficulty.
Solution Approach 2:
The patent introduces edge region patterns as an intermediary reference standard. By first establishing correct matching relationships at the edge region where patterns are less deformed, the system creates a verification benchmark that can be used to judge the correctness of central region matching, even when pitch shifts occur.
2Productivity
If CD measurement is performed using only pattern profile features to simplify the process, then measurement speed is improved, but the ability to analyze design or process problems is lost
Solution Approach 1:
The patent performs preliminary classification of patterns into first and second groups based on layer information before measurement. This preliminary organization enables the system to not only measure CD values but also associate them with specific design layers and processes, thereby preserving design and process information while maintaining efficient measurement through automated classification and integration.
3Productivity
If multi-patterning is used to form narrower pattern intervals to achieve higher integration, then circuit integration density is improved, but pattern dimension variation increases making CD measurement difficult
Solution Approach 1:
The patent applies different matching and measurement strategies to different regions (edge region vs. central region) and different pattern groups (first and second groups based on layer information). This local differentiation allows the system to account for and compensate for pattern dimension variations that occur in multi-patterning processes, maintaining measurement accuracy despite increased integration density.
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
Ensures accurate pattern matching and allows for optimizing process parameters by monitoring pattern widths in areas with small process margins, thereby improving the precision of multi-patterning processes.
Implementation Method 1
generating an image of a pattern on the workpiece by electron-beam irradiation
Data Source
AI summary
The method includes: determine a first integrated value by integrating measured values of widths of reference patterns (210A) belonging to a first group; determine a second integrated value by integrating measured values of widths of reference patterns (210B) belonging to a second group; performing second matching between patterns on an image of a second region and corresponding CAD patterns; determining a third integrated value by integrating measured values of widths of patterns (220A) belonging to a first group; determining a fourth integrated value by integrating measured values of widths of patterns (220B) belonging to a second group; and determining that the second matching has been performed correctly when the magnitude relationship between the third integrated value and the fourth integrated value coincides with the magnitude relationship between the first integrated value and the second integrated value.