Substrate Pre-Alignment Using Multi-Camera Fiducial and Bow Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional substrate pre-aligners struggle to accurately align bonded wafers, particularly those with fiducials obscured by bonding agents like wax, and fail to effectively determine substrate bow or warp.
Innovation Solution
A substrate pre-aligner system utilizing multiple cameras and a machine-learning framework to capture and process images of the substrate, allowing for the detection of fiducials, calculation of alignment offsets, and determination of substrate bow or warp.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional pre-aligners are used to align substrates, then the alignment process is simple, but the alignment precision deteriorates when fiducials are obscured by bonding agents
Solution Approach 1:
The system segments the alignment task into multiple components: fiducial detection, bow measurement, and alignment calculation are performed separately using different image processing techniques. This allows each component to be optimized independently, maintaining high precision even when fiducials are partially obscured.
Solution Approach 2:
The system transitions from 2D image analysis to 3D surface reconstruction by capturing images at multiple focal depths. This enables the system to measure substrate bow and compensate for out-of-plane deviations, significantly improving alignment precision for bonded wafers with significant curvature.
2Measurement precision
If multiple cameras are used to capture unique views of the substrate, then the measurement capability improves, but the device complexity increases
Solution Approach 1:
The multi-camera system is designed to perform multiple functions simultaneously: fiducial detection, edge detection, bow measurement, and alignment calculation. By making the system universal, the added complexity is justified by the significant improvement in measurement capabilities and the elimination of separate measurement devices.
Solution Approach 2:
The system merges fiducial detection, edge detection, and bow measurement into a single integrated image processing workflow. Multiple camera views are combined and processed together to generate comprehensive substrate characterization data, reducing the need for separate measurement steps and simplifying the overall process.
3Reliability
If conventional image processing is used, then the processing speed is fast, but the ability to detect obscured fiducials deteriorates
Solution Approach 1:
The system introduces intermediate processing steps including multi-scale image analysis and adaptive thresholding that act as mediators between raw images and fiducial detection. These intermediate steps enhance the visibility of obscured fiducials by processing images at different scales and applying adaptive algorithms, improving detection reliability without excessive time penalty.
Solution Approach 2:
The system performs preliminary image processing operations such as noise filtering, contrast enhancement, and edge detection before attempting fiducial identification. By preparing the images in advance with these preliminary actions, the subsequent fiducial detection becomes more reliable even for obscured features, and the overall process is optimized to minimize total processing time.
Data Source
AI summary
Various examples include a substrate pre-aligner system that can align substrates by detecting a fiducial on the substrate, determine an amount of bow in the substrate, and determine other characteristics of the substrate. In one example, by imaging both a 0-degree orientation and after a single 180-degree rotation of the substrate, the pre-aligner of the disclosed subject-matter can determine, for example, a location of the fiducial and bow in the substrate. In other embodiments, multiple cameras are used to capture images of the substrate substantially simultaneously and determine, for example, a location of the fiducial and bow in the substrate. The multiple camera embodiment can also allow a higher throughput of substrates as compared with the 0-degree to 180-degree embodiment. Other systems and methods are also disclosed.