Wafer Interface Vision Inspection for Electrodeposition Tool Uptime
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing wafer processing tools face challenges in detecting and addressing conditions such as wetness, dirtiness, damage, and inconsistent electrical contact at the wafer interface, which can lead to defective wafers and reduced product yield.
Innovation Solution
The implementation of a machine learning classifier, specifically a residual neural network, in conjunction with a camera system and logic machine, to acquire images of the wafer interface and classify them as 'normal', 'wet', 'dirty', 'damaged', or 'ambiguous', allowing for automatic triggering of maintenance programs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional inspection methods are used for wafer interface conditions, then manual detection is possible, but inspection frequency is low and tool downtime increases
Solution Approach 1:
The patent replaces manual mechanical inspection with an automated optical inspection system using a camera and machine learning classifier. The camera captures images of the wafer interface, and the machine learning model automatically classifies conditions such as wetness, dirtiness, and damage, enabling frequent inspections without manual intervention and reducing tool downtime.
2Measurement precision
If manual inspection of wafer interface is performed, then detection is possible, but detection precision and early issue identification are limited
Solution Approach 1:
The patent replaces manual visual inspection with an automated machine learning-based classification system. The camera captures detailed images of the wafer interface, and the trained machine learning model accurately classifies various conditions including wetness, dirtiness, and damage, providing superior detection precision compared to manual methods and enabling early issue identification to maintain wafer yield.
3Reliability
If frequent inspections are implemented using manual methods, then more issues can be detected, but labor requirements and operational complexity increase
Solution Approach 1:
The patent replaces complex manual inspection procedures with a streamlined automated system consisting of a camera and machine learning classifier. This substitution maintains high reliability for issue detection while reducing operational complexity, as the automated system requires no manual labor and can perform frequent inspections independently.
Solution Approach 2:
The inspection system is self-service in nature, automatically capturing images and classifying wafer interface conditions without requiring human operators. The machine learning model independently processes images and generates classifications, enabling the system to perform frequent inspections autonomously and reducing the complexity associated with manual inspection workflows.
Data Source
AI summary
Examples are disclosed that relate to diagnosing a condition of a wafer processing tool using a machine learning classifier. One example provides an electrodeposition tool comprising a cup. The cup comprises a wafer interface. The wafer interface comprises a lip seal and a plurality of electrical contacts. The electrodeposition tool further comprises a camera positioned to image at least a portion of the wafer interface. The electrodeposition tool further comprises a logic machine, and a storage machine storing instructions executable by the logic machine. The instructions are executable to acquire an image of the wafer interface via the camera. The instructions are further executable to obtain a classification of the image of the wafer interface from a trained machine learning function. The instructions are further executable to control the electrodeposition tool to take an action based on the classification.