Weighted Regression Thickness Map for CMP Endpoint Control
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, leading to inconsistencies in achieving desired flatness or thickness, which existing optical monitoring techniques fail to adequately address.
Innovation Solution
An in-situ monitoring system measures spectra at multiple positions on the substrate, generating a wafer-level thickness map by fitting optical models and weighting values based on goodness of fit to reference spectra, allowing for adjustments in polishing parameters to improve within-wafer and wafer-to-wafer uniformity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If in-situ optical monitoring is used to measure thickness during polishing, then real-time feedback for endpoint detection is achieved, but measurement precision deteriorates due to large spot size and high relative motion between probe and substrate
Solution Approach 1:
The patent introduces a reference spectrum as an intermediary standard for comparison. By comparing measured spectra against a reference spectrum obtained under ideal conditions, the system can compensate for measurement variations caused by large spot size and high relative motion, thereby maintaining measurement precision while enabling real-time monitoring during polishing
Solution Approach 2:
The patent applies regression analysis to transform raw spectral measurements into accurate thickness values. By using multiple measured spectra and applying statistical regression techniques, the system converts imprecise individual measurements into a reliable thickness map, resolving the precision issue while maintaining real-time capability
2Productivity
If spectral measurements are taken with large spot size and high relative motion for wafer-level mapping, then throughput is improved, but measurement precision deteriorates
Solution Approach 1:
The patent divides the substrate into multiple measurement positions and collects spectra at each position. By segmenting the wafer into a grid of measurement points and collecting spectra at each point, the system achieves comprehensive wafer-level coverage while maintaining manageable data processing requirements, thus improving throughput without sacrificing overall mapping accuracy
Solution Approach 2:
The patent uses regression analysis to process multiple measured spectra at each position and generate a weighted thickness value. This feedback mechanism processes multiple measurements to produce a reliable thickness map, allowing the system to use larger spot sizes and faster scanning speeds while maintaining precision through statistical processing of the collected spectral data
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 provides a more accurate thickness map, reduces within-wafer and wafer-to-wafer non-uniformity, and enhances the reliability of detecting the polishing endpoint, ensuring consistent substrate planarization.
Implementation Method 1
measuring a plurality of spectra reflected from a substrate at a plurality of different positions on the substrate
Data Source
AI summary
A method of controlling a polishing operation includes measuring a plurality of spectra at a plurality of different positions on a substrate to provide a plurality of measured spectra. For each measured spectrum of the plurality of measured spectra, a characterizing value is generated based on the measured spectrum. For each characterizing value, a goodness of fit of the measured spectrum to another spectrum used in generating the characterizing value is determined. A wafer-level characterizing value map is generated by applying a regression to the plurality of characterizing values with the plurality of goodnesses of fit used as weighting factors in the regression. A polishing endpoint or a polishing parameter of the polishing apparatus is adjusted based on the wafer-level characterizing map, and the substrate or a subsequent substrate is polished in the polishing apparatus with the adjusted polishing endpoint or polishing parameter.