3D Profiling Semiconductor Chips via SEM Image PCA
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing miniaturization of semiconductor chips requires precise 3D profiling techniques for measuring etching depth and pitch dimensions, but existing methods using scanning electron microscopy (SEM) are limited to 2D imaging, necessitating destructive structural analysis for obtaining 3D information.
Innovation Solution
A 3D profiling system that utilizes SEM images and gray level information, performing principal component analysis (PCA) and multiple linear regression to generate 3D profiles of semiconductor chips, allowing for non-destructive, precise measurement of fine patterns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If SEM is used for structural analysis of fine patterns, then non-destructive measurement is achieved, but only 2D images are obtained without 3D information
Solution Approach 1:
The patent transforms 2D SEM image data into 3D profile information by introducing a depth dimension through statistical analysis. Multiple SEM images taken at different focus settings are processed to extract height information, converting planar images into three-dimensional structural data without physical sectioning
2Measurement precision
If destructive structural analysis is used to obtain 3D information, then accurate 3D profile is achieved, but the sample is damaged and measurement time increases
Solution Approach 1:
The patent creates virtual 3D models from multiple 2D SEM images through statistical reconstruction. Instead of physically sectioning and examining samples at different depths, the system generates computational copies representing cross-sectional profiles at various depths, dramatically reducing measurement time while maintaining accuracy
3Ease of operation
If traditional 2D SEM imaging is used, then measurement process is simple, but 3D structural information cannot be obtained
Solution Approach 1:
The patent changes the interpretation parameters of SEM images by analyzing gray level variations and applying statistical transformations. By processing images with different focus settings and applying principal component analysis, the system extracts height information from intensity variations, converting standard 2D imaging parameters into 3D structural 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
Enables rapid, non-destructive 3D structural analysis of semiconductor chips, improving processing efficiency and reducing data processing time by extracting principal components from SEM images, thereby enhancing the accuracy of 3D profiling.
Implementation Method 1
an electron irradiation unit that irradiates an electron beam toward semiconductor devices on the chip
Implementation Method 2
a detecting unit that detects intensity of the electrons emitted from the semiconductor devices to generate a gray level of an SEM image
Data Source
AI summary
A 3D profiling system of a semiconductor chip is provided and includes a storage unit that receives scanning electron microscope (SEM) images of a plurality of semiconductor devices having respective data with respect to a plurality of different components and gray levels of each SEM image. An extraction unit that performs principal component analysis (PCA) on the gray level of the SEM image and separates principal components from among the plurality of different components is also part of the system. Additionally, a calculation unit receives provision of actually measured values of the plurality of semiconductor devices, and applies a multiple linear regression to the principal components based on the measured values to complete a 3D profile of the semiconductor chip.