CD-SEM Pattern Dimension Measurement Using Simulated Waveform Library
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for measuring semiconductor patterns using CD-SEM are prone to measurement errors due to variations in SEM signal waveforms caused by pattern shape changes, and estimating apparatus parameters is time-consuming and prone to inconsistencies.
Innovation Solution
The method employs electron beam simulation (Monte Carlo simulation) to create SEM images that reflect apparatus characteristics, storing simulated waveforms in a library for high-precision measurements by selecting the most similar waveform to the actual image, thereby reducing parameter estimation and improving accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If model-based measurement methods are used to achieve high-precision measurements, then measurement precision is improved, but apparatus parameter estimation takes a large amount of processing time and may produce inconsistent results
Solution Approach 1:
The patent pre-calculates and stores simulation waveforms for various pattern shapes in a library before actual measurements are performed. This preliminary action eliminates the need for time-consuming parameter estimation during measurement, as the system directly compares actual waveforms against pre-stored simulation results to determine pattern dimensions.
2Adaptability or versatility
If apparatus parameters are estimated during matching procedure, then measurement adaptability is improved, but measurement consistency deteriorates due to inconsistent estimation results
Solution Approach 1:
The patent creates a library of pre-calculated simulation waveforms that copy the expected signal responses for various pattern shapes. Instead of estimating parameters during measurement, the system copies the closest matching waveform from the library to determine pattern dimensions, ensuring consistent and reproducible results across different measurements.
3Productivity
If conventional measurement methods are used, then measurement speed is improved, but measurement precision deteriorates due to SEM signal waveform changes with pattern shape
Solution Approach 1:
The patent pre-calculates simulation waveforms for various pattern shapes and stores them in a library before measurements are performed. During actual measurement, the system quickly compares the measured waveform against the pre-stored library to determine pattern dimensions, achieving both high speed and high precision without requiring time-consuming parameter estimation during measurement.
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 enables consistent, high-speed, and high-precision measurements by predetermining apparatus parameters, reducing measurement variations and calculation time, and enhancing electron beam simulation accuracy.
Implementation Method 1
performing an electron beam simulation (Monte Carlo simulation) by providing apparatus characteristic information of an electron microscope
Data Source
AI summary
To provide a consistent, high-speed, high-precision measurement method based on an electron beam simulation by reflecting the apparatus characteristics of a CD-SEM in an electron beam simulation, the present invention discloses a method for measuring a measurement target pattern with a CD-SEM, the method comprising the steps of performing an electron beam simulation on various target pattern shapes, which is reflected apparatus characteristic and image acquisition conditions; creating SEM simulated waveforms; storing a combination of the created SEM simulated waveforms and pattern shape information corresponding to the created SEM simulated waveforms as a library; comparing an acquired actual electron microscope image with the SEM simulated waveforms; selecting the SEM simulated waveform that is most similar to the actual electron microscope image; and estimating the shape of the measurement target pattern from the pattern shape information corresponding to the selected SEM simulated waveform.