2D Semiconductor Edge Roughness Estimation Using PSD Denoising
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for roughness estimation in semiconductor fabrication are limited to one-dimensional features and suffer from noise interference, leading to inaccurate measurements of edge roughness in two-dimensional features, which can impact device performance and yield.
Innovation Solution
A computerized system that corrects actual contours of two-dimensional features using transformations relative to reference contours and applies power spectral density analysis to generate accurate PSD data, followed by a denoising process to remove segmentation noise.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional roughness estimation methods are applied to two-dimensional features, then the measurement process is simple, but the measurement precision deteriorates due to noise interference and inability to handle 2D geometry
Solution Approach 1:
The patent segments the edge profile into discrete contour points and divides the frequency spectrum into distinct bands (low frequency, mid frequency, high frequency). This segmentation allows independent analysis of different roughness components and enables the application of targeted correction algorithms to each segment, improving overall measurement precision while managing system complexity through modular processing.
Solution Approach 2:
The patent introduces an intermediary correction model that transforms the raw contour data into a corrected contour by removing transformation artifacts. This intermediary step acts as a mediator between the raw measurement and the final roughness calculation, eliminating noise interference and enabling accurate 2D feature analysis without requiring complex direct measurement methods.
2Measurement precision
If contour transformation correction is applied to align actual contours with reference contours, then the measurement precision improves by reducing artifacts, but the processing time increases
Solution Approach 1:
The patent performs preliminary contour transformation correction before the final roughness measurement. By pre-aligning the actual contours with reference contours and removing transformation artifacts in advance, the subsequent roughness calculation can proceed more efficiently with cleaner data, reducing overall processing time while improving precision.
Solution Approach 2:
The patent changes the parameter representation from raw pixel coordinates to corrected contour coordinates through transformation matrices. This parameter change allows the system to work with normalized, artifact-free data throughout the analysis process, reducing computational complexity and processing time while maintaining high measurement precision.
3Adaptability or versatility
If power spectral density analysis is applied to 2D features, then the adaptability improves for analyzing complex geometries, but the difficulty of detecting and measuring increases due to noise and segmentation artifacts
Solution Approach 1:
The patent extracts and removes segmentation artifacts and noise from the contour data before applying PSD analysis. By taking out the harmful components (artifacts, noise) separately from the useful signal (true edge roughness), the system can accurately analyze 2D features without the interference of measurement errors, maintaining adaptability while reducing measurement difficulty.
Solution Approach 2:
The patent transitions from spatial domain analysis to frequency domain analysis through PSD transformation. This dimensional change allows the system to separate different types of noise and artifacts by frequency, making it easier to detect and remove unwanted signals while preserving the true roughness characteristics of 2D features, thus reducing measurement difficulty.
4Ease of manufacture
If segmentation noise is present in the contour data, then the ease of manufacture improves by using standard imaging processes, but the measurement precision deteriorates due to noise interference in PSD data
Solution Approach 1:
The patent converts the harmful segmentation noise into a detectable pattern by applying PSD analysis. The noise, while detrimental to measurement precision, creates a characteristic frequency signature that can be identified and separated from the true roughness signal. By transforming the noise into the frequency domain, the system can actually utilize the noise characteristics to guide the correction process, turning a harmful factor into a beneficial diagnostic tool.
Data Source
AI summary
There is provided a system and method of estimating edge roughness of a feature on a semiconductor specimen. The method includes obtaining a set of images capturing the feature and design data of the feature; providing, for each given image in the set, a target contour of the feature in the given image, giving rise to a set of target contours corresponding to the set of images, wherein the target contour is obtained by correcting an actual contour of the feature extracted from the given image, with respect to a transformation between the actual contour and a reference contour of the feature obtained from the design data; and generating power spectral density (PSD) data based on edge placement difference (EPD) between each target contour in the set of target contours and the reference contour, wherein the PSD data is usable for estimating edge roughness of the feature.