Acoustic Endpoint Detection for Semiconductor Grinding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for detecting the end point of grinding in semiconductor manufacturing are not accurate, particularly when processing a surface with a target layer on a stopper layer, as they rely on inefficient detection methods that can delay or miss the endpoint.
Innovation Solution
A control device and method that collect and analyze the sound of processing using a sound sensor, calculating the difference in power spectrum between two points in time to determine the change point of processing, allowing for precise control of the grinding device to stop at the correct endpoint.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional detection methods are used for grinding endpoint detection, then the detection method is simple to implement, but the detection accuracy is low and endpoint detection may be delayed or missed
Solution Approach 1:
The patent replaces conventional mechanical or optical endpoint detection methods with acoustic field-based detection. A sound sensor collects grinding sounds during the grinding process, and the system analyzes acoustic signals to detect the endpoint. This substitution of detection mechanism achieves higher accuracy while maintaining reasonable system complexity.
Solution Approach 2:
The patent transforms the detection approach by changing from direct mechanical/optical measurement to acoustic parameter analysis. The system extracts features from the acoustic signal spectrum and analyzes temporal changes in these parameters to identify the grinding endpoint, achieving improved detection accuracy through parameter transformation.
2Measurement precision
If acoustic signal analysis is used for endpoint detection, then the endpoint detection accuracy is improved, but the processing time and computational complexity increase
Solution Approach 1:
The patent extracts only the essential acoustic features from the full acoustic signal spectrum that are relevant for endpoint detection. By identifying and extracting key spectral characteristics rather than analyzing the entire signal, the system achieves accurate endpoint detection while reducing computational load and processing time.
Solution Approach 2:
The system performs preliminary analysis of the acoustic signal characteristics during the grinding process, continuously monitoring and preparing the data for endpoint detection. This preliminary action allows the system to quickly identify when endpoint conditions are met, reducing the time from signal generation to endpoint determination.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enables accurate and timely detection of the grinding endpoint, reducing the likelihood of delayed or missed endpoint detection, and ensuring reliable stopping of the grinding process.
Implementation Method 1
a sound sensor 18 which collects a grinding sound of the wafer 11
Data Source
AI summary
A control device of a semiconductor manufacturing apparatus includes a processor and a memory connected to the processor and storing instructions executable by the processor. The instructions collect a sound of processing a substrate by the semiconductor manufacturing apparatus. The instructions calculate a difference of a power spectrum of the processing sound between a first point of time and a second point of time. The instructions determine a change point of processing of the substrate based on the difference.