Automated EDX Review for Semiconductor Defect Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional automated inspection systems for semiconductor wafers and similar substrates face inefficiencies in defect classification and elemental composition analysis, requiring manual intervention and lengthy processes, which hinder real-time decision-making and yield improvement.
Innovation Solution
An automated method and apparatus that utilize energy-dispersive x-ray (EDX) spectroscopy in conjunction with scanning electron microscopy (SEM) for defect classification and elemental analysis, allowing for automated selection and review of specific defect types, with improved reference site alignment for more accurate elemental information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual defect classification and EDX review processes are used, then accurate defect analysis can be achieved, but the time and effort required increases significantly
Solution Approach 1:
The system performs automated defect classification and EDX review without manual intervention. The automated classification algorithm independently categorizes defects, and the system automatically selects appropriate defects for EDX analysis and aligns reference sites, eliminating the need for manual review while maintaining accuracy
Solution Approach 2:
Manual mechanical classification and review processes are replaced with automated computational algorithms. The system uses image processing algorithms for defect classification and automated control systems for EDX acquisition and reference site alignment, substituting human operators with automated electronic systems
2Measurement precision
If comprehensive EDX review is performed on all defects, then complete elemental analysis is achieved, but the process becomes excessively lengthy and inefficient
Solution Approach 1:
The system extracts and performs EDX analysis only on specific defect types that require elemental composition information, rather than conducting comprehensive EDX review on all defects. The automated classification system identifies which defects need EDX analysis, extracting only the necessary cases for detailed elemental analysis
Solution Approach 2:
Instead of performing complete EDX analysis on every defect, the system applies partial action by selecting only certain defect types for EDX review based on classification results. This selective approach provides sufficient elemental analysis information while significantly reducing the overall analysis time
3Productivity
If automated classification is implemented, then review efficiency improves, but the complexity of the system increases
Solution Approach 1:
The automated system segments the defect review process into distinct functional modules: defect detection, automated classification, EDX selection, and reference site alignment. Each module performs a specific function independently, making the overall complex system manageable through functional segmentation
4Measurement precision
If manual reference site alignment is performed, then accurate elemental information can be obtained, but the time and effort required increases
Solution Approach 1:
Manual reference site alignment operations are replaced with automated image processing and alignment algorithms. The system automatically processes defect images, identifies corresponding reference sites, and performs alignment without manual intervention, maintaining accuracy while simplifying operation
Solution Approach 2:
The system performs self-alignment of reference sites using automated algorithms that independently analyze defect images and determine proper alignment without requiring operator intervention. The automated system serves itself by automatically completing the alignment function
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 reduces the time and effort required for EDX review, providing more accurate elemental and morphological data for defect classification, enabling faster and more precise identification of defect types and their causes, thus improving manufacturing yield.
Implementation Method 1
an electron beam striking a sample in an SEM also produces x-rays that are characteristic of the material of the sample
Implementation Method 2
a solid state detector is positioned relatively close to the sample to collect x-rays emanating from the sample due to impingement by the electron beam
Implementation Method 3
a manufactured substrate (such as a silicon wafer or a reticle) is scanned with a focused beam of electrons which results in the emission of secondary electrons from the substrate surface
Data Source
AI summary
One embodiment relates to a method for automated review of defects detected in a defective die on the target substrate. The method includes: performing an automated review of the defects using an secondary electron microscope (SEM) so as to obtain electron-beam images of the defects; performing an automated classification of the defects into types based on morphology of the defects as determined from the electron-beam images; selecting defects of a specific type for automated energy-dispersive x-ray (EDX) review; and performing the automated EDX review on the defects of the specific type. In addition, automated techniques are disclosed for obtaining an accurate reference so as to improve the usefulness of the EDX results. Furthermore, an automated method of classifying the defects based on the EDX results is disclosed which provides a final pareto that combines both morphological and elemental information. Other embodiments, aspects and features are also disclosed.