Wafer Grinding Control Using Predicted Nanotopography
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current wafer manufacturing processes, particularly double-sided grinding, often result in topology degradation due to misalignment of hydrostatic pads and grinding wheels, leading to nanotopography defects like C-Marks and B-Rings, which cause significant yield losses and are not efficiently detected until after processing hundreds of wafers.
Innovation Solution
A computer device and method that analyze wafer surfaces using a model to predict post-processing nanotopography, allowing for real-time adjustments to the grinder to prevent topology degradation by comparing scan data to thresholds and adjusting the grinding process accordingly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional inspection methods are used to detect nanotopography defects, then detection capability is limited, but significant time and material are lost because problems are detected only after processing hundreds of wafers
Solution Approach 1:
The system performs preliminary nanotopography measurement and analysis before the wafer undergoes extensive processing. By detecting potential defects early in the process chain, the system can identify problematic wafers before significant time and material are invested in downstream processing, thus reducing waste while maintaining high detection precision through advanced measurement techniques.
Solution Approach 2:
The system replaces traditional mechanical contact-based inspection methods with non-contact optical measurement techniques for nanotopography detection. This substitution enables higher measurement precision and earlier defect detection without the physical interference and time constraints of mechanical systems, allowing problems to be identified before hundreds of wafers are processed.
Data Source
AI summary
A computer device includes at least one processor in communication with at least one memory device. The at least one processor is programmed to store, in the at least one memory device, a model for simulating a portion of an assembly line and receive scan data of a first inspection of a product being assembled, execute the model using the scan data as inputs to generate a final profile of the product, compare the final profile to one or more thresholds, determine if the final profile exceeds at least one of the one or more thresholds, and adjust the first device if the determination is that the final profile exceeds at least one of the one or more thresholds.