Non-destructive 3D Profiling of Semiconductor Features via SEM Energy Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current semiconductor manufacturing processes face challenges in accurately characterizing three-dimensional profiles of high, mid, and low aspect ratio structures due to the limitations of destructive imaging techniques like SEM and TEM, which provide limited statistical data and are time-consuming, prone to variability, and not suitable for large-scale, non-destructive wafer analysis.
Innovation Solution
A method using scanning electron microscopy (SEM) with energy filter values to obtain two-dimensional planar geometrical profile images at various depths, simulating potential distribution within the structure, and converting these images into composite three-dimensional profiles along the longitudinal axis, enabling non-destructive, high-resolution, and cost-effective characterization of semiconductor features.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If destructive imaging techniques like x-SEM or TEM are used to accurately reveal 3D profiles, then measurement precision is improved, but productivity deteriorates due to time-consuming sample preparation and limited wafer throughput
Solution Approach 1:
The patent creates virtual copies of the semiconductor structure at different depths by simulating electron transport and generating synthetic cross-sectional images. Instead of physically sectioning the wafer, the system computationally reconstructs multiple 2D slices from a single non-destructive top-down SEM image, providing full 3D profile information without sample preparation or wafer destruction.
Solution Approach 2:
The patent replaces the mechanical sample preparation process (physical sectioning with diamond saws, mounting, and polishing) with a computational approach. By solving the electron transport equation and simulating electron scattering, the system substitutes physical destruction with mathematical modeling to achieve the same diagnostic information.
2Manufacturing precision
If destructive cross-sectional imaging is performed to obtain accurate 3D profiles, then manufacturing precision is improved, but loss of time increases due to semi-manual sample preparation
Solution Approach 1:
The system performs self-characterization by using the same top-down SEM imaging data to generate both the depth map and the cross-sectional profiles. The electron transport simulation automatically extracts structural information without requiring external sample preparation steps, making the process self-sufficient and eliminating time-consuming manual operations.
Solution Approach 2:
The patent performs preliminary computational preparation by pre-calculating the electron transport simulation and potential distribution before generating cross-sectional images. This pre-processing creates a library of synthetic images at different depths, allowing rapid extraction of 3D profile information without time-consuming physical sample preparation when measurements are needed.
3Ease of operation
If conventional SEM imaging is used for wafer inspection, then ease of operation is maintained, but measurement precision deteriorates due to inability to obtain longitudinal cross-section information
Solution Approach 1:
The patent makes the top-down SEM imaging system multi-functional by enabling it to simultaneously provide both planar topography information and longitudinal cross-sectional profiles. The same imaging setup and electron transport simulation that reveal surface features also extract depth information and generate virtual cross-sections, eliminating the need for specialized cross-sectional SEM equipment while maintaining operational simplicity.
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 allows for accurate, non-destructive, and automated characterization of semiconductor features across the entire wafer, improving throughput and reducing costs by integrating with production lines, providing higher accuracy and full-wafer coverage for various aspect ratio structures.
Implementation Method 1
obtaining a plurality of energy filter values using a model that simulates potential distribution within the 3D feature when an electron beam of the SEM impinges on the selected area including the 3D feature
Implementation Method 2
extracting a correspondence between the plurality of energy filter values and respective depths of the 3D feature along a longitudinal direction by analyzing the simulated potential distribution
Data Source
AI summary
A plurality of energy filter values are obtained using a model that simulates potential distribution within a 3D feature when an electron beam of an SEM impinges on a selected area that includes the 3D feature. A correspondence is extracted between the plurality of energy filter values and respective depths of the 3D feature along a longitudinal direction by analyzing the simulated potential distribution. A plurality of SEM images of the 3D feature corresponding to the plurality of energy filter values are obtained. The plurality of SEM images are associated with their respective depths based on the extracted correspondence between the plurality of energy filter values and the respective depths. A composite 3D profile of the 3D feature is generated from the plurality of SEM images obtained from various depths of the 3D feature.