Optical Emission Inspection for Reliable Plasma Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing substrate processing methods face challenges in ensuring reliable and efficient plasma processing due to issues such as outdoor air introduction and gas mixing, which can affect the quality and consistency of semiconductor manufacturing processes.
Innovation Solution
An inspection method utilizing optical emission spectrometry to collect and analyze optical signals, applying a fitting function to scale and convert data into peak probabilities, and comparing these with reference data to identify normal conditions, enabling reliable plasma processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If optical emission spectrometry is used to collect and analyze optical signals with complex data processing, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The spectrum data is divided into multiple sections (first section, second section, etc.) along the wavelength axis. Each section is processed independently to determine peak probabilities, which are then combined for final analysis. This segmentation allows complex spectral analysis to be broken down into manageable segments while maintaining high detection precision.
Solution Approach 2:
A fitting function f(x)=Ae^(-ω|x|^n)+c is introduced as an intermediary to model and analyze the spectrum data. This mathematical function serves as a mediator between the raw optical signals and the final detection results, enabling precise identification of gas species and their concentrations without requiring complex direct analysis of the raw data.
2Reliability
If comprehensive data analysis with fitting functions is applied, then reliability is improved, but loss of time increases
Solution Approach 1:
The fitting function f(x)=Ae^(-ω|x|^n)+c is pre-established and optimized before actual gas detection. The function parameters and structure are determined in advance through theoretical analysis and calibration, so that during real-time inspection, only simple parameter fitting is needed rather than complex analysis, significantly reducing inspection time while maintaining high reliability.
Solution Approach 2:
The fitting function uses adjustable parameters (A, ω, n, c) that can be optimized based on different gas species and detection conditions. By changing these parameters according to the specific detection task, the system achieves high reliability for identifying different gases without requiring completely different analysis methods for each case.
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
Enhances the reliability and performance of substrate processing by accurately detecting gas introduction and mixing, ensuring consistent plasma conditions and improving the quality of semiconductor manufacturing processes.
Implementation Method 1
collecting spectrum data of a first optical signal in a chamber by using optical emission spectrometry
Implementation Method 2
an electromagnetic field is formed in an inner space of a chamber, and the electromagnetic field excites a process gas provided into the chamber to a plasma state
Implementation Method 3
an electromagnetic field is formed in an inner space of a chamber, and the electromagnetic field excites a process gas provided into the chamber to a plasma state
Data Source
AI summary
An inspection method includes extracting a first similarity by comparing first data of a first optical signal with reference data of a reference optical signal, generating a first normal distribution of the first similarity, extracting a second similarity by comparing second data of a second optical signal with the reference data of the reference optical signal, generating a second normal distribution of the second similarity, and comparing the first normal distribution with the second normal distribution. The extracting of the first similarity includes deriving the first data of the first optical signal.