Charged Particle Beam Device Pattern Search Automation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current charged particle beam devices face challenges in maintaining high accuracy and throughput during semiconductor measurement and inspection, as they require time-consuming recipe optimization and are prone to addressing errors, especially when pre-registered templates are inappropriate or missing.
Innovation Solution
A charged particle beam device with an arithmetic processing unit that generates images from detection signals, searches for patterns using templates, and adjusts the beam position based on deviations between detected and pre-registered positions, allowing for pattern search and field of view alignment without pre-registration, and uses wide region data to suppress addressing errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If a pre-registered template is used for pattern search, then the measurement process can be automated, but addressing errors occur when the template is inappropriate or missing
Solution Approach 1:
The system performs preliminary actions by acquiring wide region data before the actual measurement process. This wide region data serves as a backup reference that can be used when pre-registered templates are inappropriate or missing, preventing addressing errors while maintaining automation.
Solution Approach 2:
The system prepares compensatory measures in advance by storing wide region data that covers areas beyond the immediate field of view. This cushioning data acts as a safety net that prevents addressing failures when the primary template matching fails, ensuring continuous reliable operation.
2Measurement precision
If recipe optimization is performed to improve measurement accuracy, then measurement precision increases, but processing time increases
Solution Approach 1:
The system performs preliminary acquisition of wide region data and prepares reference images before the actual measurement process. By having these reference materials ready in advance, the system eliminates the need for time-consuming recipe optimization during measurement, achieving both high precision and fast processing.
Solution Approach 2:
The system acquires wide region data that extends beyond the immediate measurement area. This excessive action of capturing more data than strictly necessary for the current measurement provides a rich reference set that enables accurate template matching without requiring iterative optimization, thus saving time.
3Productivity
If the field of view is limited to the measurement area, then the measurement process is efficient, but addressing errors occur when patterns are outside the current view
Solution Approach 1:
The system transitions from a single-field-of-view approach to a multi-scale approach by acquiring wide region data that encompasses multiple fields of view. This dimensional expansion allows the system to maintain efficient localized measurements while having access to broader contextual information for reliable pattern searching.
Solution Approach 2:
The system performs preliminary acquisition of wide region data that covers areas beyond the immediate measurement field. This advance preparation ensures that when pattern matching is needed, reference patterns are available even if they fall outside the current narrow field of view, maintaining both efficiency and reliability.
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 maintains a high success rate for pattern searches, reduces the frequency of recipe updates, and minimizes addressing errors, enabling efficient and accurate measurement and inspection processes.
Implementation Method 1
a charged particle beam emitted from a charged particle source is irradiated to a sample
Data Source
AI summary
Proposed is a charged particle beam device including an arithmetic processing unit that generates an image of a sample, based on a detection signal that is detected based on irradiation to the sample with a charged particle beam emitted from a charged particle source. The arithmetic processing unit searches a second image as a search target image with use of a first image as a template, and when a region corresponding to the first image is not detected in the second image, the arithmetic processing unit searches a third image that represents a region larger than a region displayed in the second image, with use of a second template.