Reference Spectra Library Generation for CMP Endpoint Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Chemical mechanical polishing (CMP) processes face challenges in achieving consistent material removal rates due to variations in substrate thickness, slurry composition, polishing pad conditions, and load, making it difficult to determine the polishing endpoint and achieve a desired profile.
Innovation Solution
A computer-implemented method is used to generate reference spectra by polishing multiple set-up substrates, measuring spectra during polishing, determining the best match among sequences of spectra, and storing these as reference spectra for optical monitoring of product substrates to adjust polishing rates and detect endpoints.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple set-up substrates are polished and spectra are measured to generate reference spectra, then the reliability and consistency of endpoint detection is improved, but the time required to begin polishing new substrates increases
Solution Approach 1:
The system performs preliminary polishing of multiple set-up substrates and generates reference spectra libraries before actual production polishing begins. This preliminary action creates ready-to-use reference data that enables fast endpoint detection during subsequent polishing operations, resolving the contradiction between thorough reference generation and quick production start-up
Solution Approach 2:
The system creates spectral copies from multiple set-up substrates to build reference libraries. Instead of measuring every production substrate against all reference spectra in real-time, the system uses pre-generated spectral copies that can be quickly compared during polishing, reducing measurement time while maintaining detection reliability
2Device complexity
If a limited number of reference spectra libraries are used, then the complexity of the system is reduced, but the ability to handle variations in polishing conditions decreases
Solution Approach 1:
The system organizes reference spectra into multiple libraries, each characterized by specific parameters such as material removal rate, substrate thickness, or polishing pad condition. This parameter-based organization allows the system to select appropriate reference libraries based on current polishing conditions, maintaining adaptability while keeping each individual library manageable in size
Solution Approach 2:
The reference spectra libraries are designed to serve multiple polishing scenarios and substrate types. Each library contains spectra that can be applied across different polishing conditions through parameter matching, allowing a limited number of libraries to handle a wide range of variations without requiring a separate library for every possible condition
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 reduces the time required to begin polishing new substrates, improves reliability and consistency of endpoint detection, and allows for efficient in-situ monitoring using a limited number of libraries, thereby reducing wafer-to-wafer non-uniformity.
Implementation Method 1
a substrate is optically monitored in-situ during polishing, e.g., through a window in the polishing pad
Data Source
AI summary
A computer-implemented method of generating reference spectra includes polishing a plurality of set-up substrates, the plurality of set-up substrates comprising at least three set-up substrates, measuring a sequence of spectra from each of the plurality of set-up substrates during polishing with an in-situ optical monitoring system to provide a plurality of sequences of spectra, generating a plurality of sequences of potential reference spectra from the plurality of sequences of spectra, determining which sequence of potential reference spectra of the plurality of sequences provides a best match to remaining sequences of the plurality of sequences, and storing the sequence of potential reference spectra determined to provide the best match as reference spectra, and selecting and storing the sequence of potential reference spectra.