CD-SEM Pattern Dimension Measurement Edge Detection Defocus
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current pattern dimension measurement methods using CD-SEM face challenges in reproducibility and accuracy due to defocus issues, which affect the measurement of fine patterns, leading to errors and difficulties in distinguishing between defocus and changes in pattern shape.
Innovation Solution
A method involving the calculation of multiple edge index positions on either side of the signal intensity maximum point to stabilize edge detection, averaging these positions to suppress fluctuations caused by defocus, and adjusting the focal point when necessary, while using dual thresholds to enhance accuracy and reduce noise effects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single edge index position is used for pattern edge detection, then the detection process is simple, but measurement precision deteriorates due to defocus and random noise causing fluctuations in edge position
Solution Approach 1:
The patent divides the edge detection process into multiple segments by calculating multiple edge index positions (first, second, third, fourth positions) at different locations relative to the signal intensity maximum point. This segmentation allows the system to capture edge information from multiple perspectives, thereby improving measurement precision by averaging out defocus and noise effects across the segmented positions.
Solution Approach 2:
The patent calculates more edge index positions than the minimum single position required for edge detection. By computing four edge index positions (exceeding the necessary one), the system obtains redundant measurement data that can be averaged to suppress fluctuations caused by defocus and random noise, thereby improving precision at the cost of increased computational complexity.
2Measurement precision
If multiple edge index positions are calculated and averaged, then measurement precision improves by suppressing defocus effects, but the detection process becomes more complex
Solution Approach 1:
The patent segments the edge detection operation into multiple discrete steps: calculating the signal intensity maximum point, determining multiple edge index positions at specific offsets from the maximum, and averaging these positions. This segmentation makes the complex process systematic and automatable, improving precision while managing operational complexity through structured procedure.
Solution Approach 2:
The patent implements a feedback mechanism where the calculated edge index positions are averaged to produce a final edge position measurement. This feedback loop allows the system to self-correct for defocus and noise effects by comparing multiple measurements and converging on an accurate result, thereby improving precision while maintaining ease of operation through automated feedback processing.
3Reliability
If conventional edge detection methods are used, then the process is simple and fast, but reliability deteriorates due to inability to distinguish defocus from pattern shape changes
Solution Approach 1:
The patent introduces a new dimension to edge detection by calculating positions at multiple offsets (positive and negative) from the signal intensity maximum point. This dimensional expansion from single-point to multi-point detection enables the system to distinguish defocus effects (which affect all positions systematically) from actual pattern shape changes, thereby improving reliability through multi-dimensional analysis.
Solution Approach 2:
The patent performs more detection operations than the minimum required by calculating four edge index positions instead of one. This excessive action provides redundant information that enables discrimination between defocus and pattern shape changes, improving reliability at the cost of increased algorithmic complexity.
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 significantly reduces measurement errors and improves reproducibility by stabilizing edge detection against defocus and random noise, ensuring accurate pattern dimension measurement even in the presence of focal point fluctuations.
Implementation Method 1
signal intensity distribution of backscattered charged particles or secondary charged particles generated from the specimen
Implementation Method 2
signal intensity distribution of backscattered charged particles or secondary charged particles generated from the specimen
Data Source
AI summary
In order to provide a pattern dimension measurement method with a small measured error and excellent reproducibility even though defocus occurs and a charged particle beam microscope used in the same, in a method for applying a charged particle beam to a specimen formed with a pattern to measure a pattern dimension from a signal intensity distribution of signal charged particles from the specimen, edge index positions (X1) and (X2) on the right and left of the maximum point of signal intensity corresponding to a pattern edge are calculated by a threshold method, and a pattern edge position (Xe) is found from a mean value between the positions. Thus, it is possible to reduce the influence of defocus on the pattern edge position (Xe).