Substrate Spectral Screening for High-Throughput Abnormality Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing substrate processing systems struggle to detect abnormalities in substrates when measured surface profiles are unavailable, as full surface profile measurements are time-consuming and impractical for high throughput.
Innovation Solution
A substrate processing system that includes a measurer to quickly measure in-plane spectral distributions of substrates, using an outlier detection model to calculate an in-plane score and detect abnormalities based on this score, even when surface profile measurements are unavailable.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If full surface profile measurements are performed to detect substrate abnormalities, then measurement precision is improved, but productivity deteriorates due to time-consuming measurements
Solution Approach 1:
The invention extracts only the necessary spectral information from the substrate surface that is sufficient for abnormality detection, rather than performing complete surface profile measurements. By using spectroscopy to capture in-plane spectral distributions at multiple positions, the system obtains critical data for detecting abnormalities while significantly reducing measurement time compared to full surface profiling.
Solution Approach 2:
The invention replaces mechanical surface profile measurement systems with optical spectroscopy-based detection. Instead of using physical probes or mechanical scanning systems to map the substrate surface, the system uses light interaction with the substrate to obtain spectral information that reveals abnormalities, thereby achieving fast, non-contact measurement that maintains detection precision without sacrificing throughput.
2Productivity
If in-plane spectral distribution measurement is used instead of surface profile measurement, then productivity is improved through faster measurement, but measurement precision may deteriorate
Solution Approach 1:
The invention changes the measurement parameter from physical surface profile geometry to optical spectral characteristics. By measuring how different wavelengths of light interact with the substrate at various in-plane positions, the system captures information about substrate composition, thickness, and structural abnormalities. This parameter transformation enables rapid optical measurement while maintaining or even improving detection sensitivity compared to mechanical profiling.
Solution Approach 2:
The invention adds the spectral wavelength dimension to the measurement process. Instead of only measuring spatial coordinates of the substrate surface, the system measures spectral intensity across multiple wavelengths at each spatial position. This additional dimensional information provides rich data for detecting substrate abnormalities, enabling accurate detection through spectral fingerprinting of different substrate regions.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
This substrate processing system includes a load port, a processing chamber, a measurer, and a controller. The load port is configured such that a storage container containing a substrate can be connected thereto. The processing chamber is configured to implement substrate processing on the substrate. The measurer is provided on a path for transporting the substrate between the storage container and the processing chamber, and is configured to measure an in-plane spectral distribution of the substrate. The controller is configured to create an outlier detection model on the basis of the in-plane spectral distribution of the substrate measured by the measurer, use the created outlier detection model to calculate an in-plane score of the substrate from the measured in-plane spectral distribution of the substrate, and detect abnormality in the substrate on the basis of the calculated score.