Wafer Cross-Section Measurement Propagation for 3D Semiconductor Metrology
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for measuring 3D semiconductor structures are computationally intensive, require significant user effort, and are prone to errors due to the complexity of photolithography processes and the three-dimensionality of semiconductor structures, making accurate and efficient defect detection and process monitoring challenging.
Innovation Solution
A method that obtains 3D measurements from multiple 2D cross section images by indicating measurement specifications in a single 2D cross section image and automatically propagating them to further images, reducing computational complexity and user effort while enhancing accuracy and flexibility.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If destructive imaging techniques using FIB-SEM are used to obtain 3D imaging datasets, then volumetric measurements of semiconductor structures can be achieved, but the process becomes computationally intensive and challenging with respect to scalability, robustness and repeatability
Solution Approach 1:
The patent segments the 3D measurement problem into multiple 2D cross-section image analyses. Instead of processing a complete 3D volumetric dataset at once, the method divides the semiconductor structure into sequential 2D slices, measures contours in each slice independently, and reconstructs 3D measurements from these segmented 2D measurements. This segmentation reduces computational complexity while maintaining measurement accuracy.
Solution Approach 2:
The patent transforms the 3D measurement problem into a series of 2D measurement problems by analyzing cross-section images. Measurement specifications defined in one 2D cross-section are propagated to adjacent 2D cross-sections, effectively reducing the dimensionality of the computational task while preserving the ability to obtain accurate 3D measurements through systematic propagation and integration of 2D data.
2Measurement precision
If traditional manual measurement methods are used by experts, then measurements can be obtained, but the process requires significant user effort and is not scalable
Solution Approach 1:
The patent implements preliminary action by automatically generating measurement specifications from a single user-defined template in the first 2D cross-section. Once the measurement specifications are defined in the initial cross-section, the system automatically propagates these specifications to all subsequent 2D cross-sections without requiring repeated user input. This preliminary definition of measurements eliminates the need for continuous user intervention throughout the measurement process.
Solution Approach 2:
The patent uses copying by replicating the measurement specifications from the first 2D cross-section to all other 2D cross-sections. The measurement template defined by the user in one cross-section is copied and applied systematically across the entire stack of 2D images, enabling automated, consistent measurements without requiring the expert to manually define measurements for each slice.
3Measurement precision
If 3D measurements are taken from dense imaging datasets comprising thousands of images, then accurate volumetric data can be obtained, but scalability and repeatability become challenging
Solution Approach 1:
The patent implements self-service by enabling the measurement system to automatically propagate specifications and perform measurements across all 2D cross-sections without requiring continuous user intervention. The system serves itself by taking the initially defined measurement template and autonomously applying it throughout the entire 3D dataset, generating complete volumetric measurements without human assistance beyond the initial setup.
Solution Approach 2:
The patent applies parameter changes by systematically varying the 2D cross-section index while maintaining consistent measurement specifications across all slices. The method processes each 2D slice with the same measurement parameters propagated from the first slice, enabling scalable and repeatable processing of large numbers of images through consistent parameter application across varying spatial positions.
Data Source
AI summary
A method of obtaining measurements of semiconductor structures on a wafer comprises: obtaining a volumetric imaging dataset of the wafer comprising multiple 2D cross section images; obtaining contours of semiconductor structures in 2D cross section images; indicating, in a first 2D cross section image, one or more measurement specifications with respect to features of contours of semiconductor structures; propagating the indicated one or more measurement specifications in the first 2D cross section image to further 2D cross section images; and obtaining measurements of semiconductor structures by evaluating the one or more measurement specifications in the first 2D cross section image and in the further 2D cross section images.