Optical Model Fitting for CMP Endpoint Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Chemical mechanical polishing (CMP) processes face challenges in determining the polishing endpoint due to variations in material removal rates caused by initial thickness, slurry composition, polishing pad conditions, and load on the substrate, making it difficult to determine completion based solely on polishing time.
Innovation Solution
An optical monitoring system fits an optical model to measured spectra during CMP, using regression techniques to determine endpoint parameters such as layer thickness, index of refraction, and extinction coefficient, allowing for real-time adjustment of polishing pressure and endpoint detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a library of reference spectra is used to match measured spectra, then endpoint detection can be performed, but the system becomes computationally intensive and time-consuming when the library is very large
Solution Approach 1:
The patent extracts only the essential spectral features (principal components) that characterize the polishing endpoint, rather than using complete reference spectra. This reduction in data dimensionality maintains detection accuracy while dramatically reducing computational time and memory requirements.
Solution Approach 2:
The patent transforms the spectral data from its original form into a reduced parameter space using principal component analysis. By changing the representation parameters from full spectra to principal component scores, the system achieves the same detection capability with far fewer parameters to process.
2Measurement precision
If complete spectral data is analyzed for endpoint detection, then detection accuracy is maintained, but the computational complexity and data processing requirements increase
Solution Approach 1:
The patent extracts the dominant variance patterns from spectral data through principal component analysis, keeping only the most informative components. This extraction process eliminates redundant information while preserving the essential characteristics needed for accurate endpoint detection.
Solution Approach 2:
The patent projects spectral data from a high-dimensional space (many wavelength points) into a lower-dimensional space (few principal components). This dimensionality change simplifies the computational problem while maintaining the ability to distinguish between different polishing states.
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 improves the reliability of endpoint detection, reduces within-wafer and wafer-to-wafer thickness non-uniformity, and is computationally less intensive compared to other techniques, enabling more precise control of the polishing process.
Implementation Method 1
a spectrum measured in-situ, e.g., during a polishing process of CMP
Implementation Method 2
measured spectra with an in-situ optical monitoring system
Implementation Method 3
The optical model is a function with multiple parameters, e.g. the thickness, index of refraction and extinction coefficient of each layer in the stack
Implementation Method 4
the thickness, index of refraction and extinction coefficient of each layer in the stack
Data Source
AI summary
A method of controlling a polishing operation includes polishing a first layer of a substrate, during polishing, obtaining a sequence over time of measured spectra with an in-situ optical monitoring system, for each measured spectrum from the sequence of measured spectra, fitting an optical model to the measured spectrum, the fitting including finding parameters that provide a minimum difference between an output spectrum of the optical model and the measured spectrum, the parameters including an endpoint parameter and at least one non-endpoint parameter, the fitting generating a sequence of fitted endpoint parameter values, each endpoint parameter value of the sequence associated with one of the spectra of the sequence of measured spectra, and determining at least one of a polishing endpoint or an adjustment of a pressure to the substrate from the sequence of fitted endpoint parameter values.