Seed Value Refinement for Scatterometric Parameter Determination
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for determining parameters of interest in lithographic processes, such as critical dimension, are hindered by the exponential increase in time and iterations required with the number of parameters, limited accuracy due to signal-to-noise ratio and Jacobian matrix condition number, and poor initial guess accuracy.
Innovation Solution
A method that uses a seed value derived from initial measurement data to assist in processing subsequent measurement data, reducing the number of iterations and improving accuracy by leveraging systematic variations across substrates and fields, and employing a model linking parameter values across multiple target patterns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If iterative modeling or library searching is used to determine parameters from scatterometer data, then measurement precision can be achieved, but the time required increases exponentially with the number of parameters
Solution Approach 1:
The patent applies preliminary action by using measurement data from a first substrate to generate seed values for parameters before processing the second substrate. This pre-computation of initial parameter estimates based on systematic variations reduces the iterative search space for subsequent measurements, thereby decreasing processing time while maintaining measurement precision
Solution Approach 2:
The patent implements feedback by using the measured parameter values from the first substrate to inform and improve the initial guesses for the second substrate. This feedback mechanism leverages systematic variations across substrates to provide better starting points for iterative modeling, reducing the number of iterations needed and thus the processing time
2Measurement precision
If the number of parameters to be determined increases, then measurement comprehensiveness improves, but the number of iterations required increases dramatically
Solution Approach 1:
The patent uses feedback from first substrate measurements to provide informed initial guesses for second substrate parameters. This feedback reduces the iteration complexity by starting closer to the solution, allowing comprehensive multi-parameter determination without dramatic increases in iteration counts
3Loss of information
If a broadband radiation beam is used in spectroscopic scatterometry, then measurement information increases, but signal-to-noise ratio decreases
Solution Approach 1:
The patent applies segmentation by dividing the broadband spectrum into multiple wavelength ranges and processing them separately. This allows optimization of signal-to-noise ratio for different spectral regions while capturing comprehensive measurement information across the entire spectrum, resolving the contradiction between information content and signal quality
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 significantly reduces the time and iterations needed to achieve accurate parameter determination, enhances measurement precision, and improves throughput in lithographic processes by using a seed value to refine subsequent measurements.
Implementation Method 1
measuring the radiation scattered therefrom
Implementation Method 2
measuring the radiation reflected therefrom
Data Source
AI summary
In a scatterometric method, different targets with different sensitivities to a parameter of interest are printed in a calibration matrix and different spectra obtained. Principal component analysis is applied to the different spectra to obtain a calibration function that is less sensitive to variation in the underlying structure than a calibration function obtained from spectra obtained from a single target.