Substrate Alignment Using Distortion-Based Mark Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing alignment methods for stacking substrates face challenges in achieving high accuracy due to the significant processing load when using all marks on the substrate, and reduced accuracy when selecting fewer marks, which is dependent on the selecting method.
Innovation Solution
An alignment method and apparatus that measures the position of marks on substrates based on distortion information, selecting marks from regions with lower distortion and higher reproducibility to improve alignment accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If all marks on the substrate are used for alignment, then alignment accuracy is improved, but processing load increases significantly
Solution Approach 1:
The patent extracts and selects only specific marks from the substrate that are least affected by distortion for alignment purposes. By identifying and using only the necessary subset of marks (those in regions with minimal distortion), the system achieves high alignment accuracy while significantly reducing the processing load compared to using all marks.
Solution Approach 2:
The patent applies local quality by evaluating distortion characteristics at different locations on the substrate and selecting marks from specific regions (those with lower distortion). This localized selection approach ensures that the chosen marks provide reliable alignment references without requiring processing of all marks across the entire substrate.
2Device complexity
If fewer marks are selected for alignment, then processing load is reduced, but alignment accuracy deteriorates
Solution Approach 1:
The patent changes the selection criterion for marks from a simple quantity-based approach to a quality-based approach using distortion information as a parameter. By evaluating and selecting marks based on their distortion characteristics rather than merely reducing the number of marks, the system maintains high alignment accuracy while reducing processing load.
3Ease of operation
If marks are selected without considering distortion information, then selection process is simplified, but alignment accuracy with other members is reduced
Solution Approach 1:
The patent performs preliminary evaluation of distortion information across the substrate before selecting marks for alignment. By pre-identifying regions with minimal distortion and selecting marks from these regions in advance, the system ensures high alignment accuracy is achieved without complicating the actual alignment process.
Data Source
AI summary
An alignment method for aligning two substrates to be stacked, comprising measuring a position of a mark selected from plurality of marks disposed on at least one substrate of the two substrates and aligning the two substrates based on the position of the measured mark, wherein the mark to be measured is selected based on information relating to distortion of the at least one substrate. The mark may be a mark disposed in a region having a smaller distortion amount of the at least one substrate than a threshold. The mark may be a mark disposed in a region having a higher reproducibility of distortion that occurs in the at least one substrate than a threshold.