Mark Position Detection Using Dynamic Region Reset
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing detection methods for alignment marks in semiconductor exposure apparatuses are prone to accuracy degradation due to dust or defects on the marks, leading to distorted one-dimensional image signals and reduced alignment precision.
Innovation Solution
A detection apparatus and method that captures two-dimensional images of marks, generates one-dimensional signals by accumulating images within a detection region, detects peaks with threshold differences, identifies and resets the detection region to exclude failure regions, and adjusts the measurement window to minimize the impact of dust or defects, thereby improving alignment accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If accumulation is performed on two-dimensional images of marks in the non-measurement direction to generate one-dimensional image signals, then the detection process is simplified and processing efficiency is improved, but measurement precision deteriorates due to distortion from dust or defects on the marks
Solution Approach 1:
The detection region in the two-dimensional image is divided into multiple regions, and accumulation is performed separately for each region. This segmentation allows the system to identify and exclude regions containing dust or defects, thereby maintaining measurement precision while preserving the efficiency benefits of accumulation processing.
Solution Approach 2:
The patent utilizes the two-dimensional image data before reducing it to one-dimensional signals. By analyzing the two-dimensional distribution of light quantities and identifying abnormal regions in both dimensions, the system can detect dust or defects that would otherwise distort the one-dimensional accumulated signal, thus maintaining measurement accuracy.
2Loss of information
If the detection region includes all areas of the mark for complete data collection, then measurement completeness is improved, but reliability deteriorates due to inclusion of dust or defect regions
Solution Approach 1:
The patent extracts and excludes regions containing dust or defects from the detection region. By identifying abnormal regions through analysis of the two-dimensional image and removing them before accumulation, the system maintains complete utilization of valid mark information while eliminating unreliable data from dust or defect areas.
Solution Approach 2:
The system performs feedback by analyzing the accumulated signal to detect abnormalities such as dust or defects, then uses this information to adjust the detection region and re-perform accumulation. This iterative feedback process ensures that only reliable regions are included in the final measurement, improving both reliability and accuracy.
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
The solution enhances the accuracy of mark position detection by reducing measurement errors caused by failure regions, ensuring precise alignment of reticle and substrate positions in semiconductor exposure processes.
Implementation Method 1
an image pickup unit for capturing an image of the mark
Implementation Method 2
an image pickup unit for capturing an image of the mark
Data Source
AI summary
A detection apparatus includes an image pickup unit and a processor which detects a position of a mark using a two-dimensional image of the mark. The processor generates a one-dimensional signal having a plurality of peaks by accumulating images included in a detection region, detects peaks in which differences between values of the peaks and a reference value are equal to or larger than a threshold value and peaks in which differences between values of the peaks and the reference value are smaller than the threshold value from among the plurality of generated peaks and obtains a failure region in the mark, resets the detection region such that the differences between the values of the detected peaks and the reference value become smaller than the threshold value, generates a one-dimensional signal by accumulating images included in the reset detection region, and detects a position of the mark.