SVM-Based X-Ray Metrology for TSV Dimension Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for monitoring the dimensions of through-silicon vias (TSVs) during semiconductor processing are either destructive or have low throughput, making it difficult to integrate nondestructive, high-throughput metrology with etching tools, which is essential for maintaining target device properties.
Innovation Solution
A method using a Support Vector Machine (SVM) machine learning model that correlates transmitted X-ray signals with TSV parameters, allowing for the quick extraction of dimensions from X-ray images during the etch process, enabling real-time process adjustments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If X-ray tomography is used to reconstruct full three-dimensional structure, then measurement precision is improved, but productivity deteriorates due to requiring very many line-of-sight X-ray images
Solution Approach 1:
The patent extracts only the essential information needed for TSV dimension measurement from the full 3D structure, using a single line-of-sight X-ray image rather than multiple images required for complete tomographic reconstruction. This extraction approach maintains measurement precision for critical parameters while eliminating the productivity penalty of acquiring numerous images.
Solution Approach 2:
The patent applies partial action by using only a single X-ray image instead of the excessive number of images required for full tomographic reconstruction. The SVM model is trained to extract sufficient dimensional information from this partial data set, achieving the necessary measurement precision without the overhead of complete 3D reconstruction.
2Measurement precision
If electron microscopy is used to inspect TSV dimensions, then measurement precision is improved, but productivity deteriorates due to process throughput reduction from redirecting samples
Solution Approach 1:
The patent replaces the mechanical sample handling and redirection process with an in-line X-ray imaging system. Instead of physically moving samples to electron microscopes, the X-ray system captures images directly at the etching tool, eliminating throughput reduction while maintaining measurement precision through advanced image analysis using SVM models.
3Productivity
If simple X-ray imaging is used instead of tomography, then productivity is improved, but measurement precision deteriorates
Solution Approach 1:
The patent changes the analytical parameters by applying SVM machine learning models to extract precise dimensional information from simple X-ray images. This parameter transformation allows the system to maintain high measurement precision for TSV dimensions while using only a single image, thus preserving productivity.
Solution Approach 2:
The patent creates a computational model (SVM) that copies and reconstructs the essential dimensional information from the simple X-ray image, effectively generating a virtual representation of TSV dimensions without requiring physical 3D reconstruction through multiple images.
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, nondestructive, and high-throughput measurement of TSV dimensions, reducing computational time to approximately 1 second, making it suitable for in-situ and in-line monitoring of etch processes, thereby maintaining target device properties.
Implementation Method 1
irradiating the structure with X-ray radiation from an X-ray source
Implementation Method 2
acquiring a transmitted X-ray radiation signal from an X-ray detector
Data Source
AI summary
Described is a method and system for measuring parameters of a structure on a substrate, such as a via or a through-silicon via (TSV) using an imaging X-ray metrology system. A previously-trained Support Vector Machine (SVM) model is used to extract structure parameters from the acquired structure X-ray images. Training of the Support Vector Machine (SVM) model is accomplished by using a library of actual or simulated X-ray images, or a combination of the two image types, paired with structure parameter sets.