Automated Defect Sampling for Semiconductor Array Regions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for creating defect samples in semiconductor manufacturing, particularly in array regions, are inefficient and time-consuming, often relying on manual processes that do not utilize available design information, leading to suboptimal sampling and increased processing time.
Innovation Solution
A system and method that incorporates design information to stack defects based on array cell types, using computer subsystems to select and generate a sample of defects, thereby improving sampling efficiency and accuracy by overlaying design data with defect positions within array cell types.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual defect sampling methods are used in array regions, then operators can select defects based on experience, but the sampling process becomes extremely time-consuming (3-5 hours) and tedious
Solution Approach 1:
The patent replaces manual mechanical sampling operations with an automated computer-based system that uses algorithms to select defects. The system automatically identifies and selects defects from array regions based on programmed criteria, eliminating the need for manual operator intervention while maintaining or improving sampling accuracy.
Solution Approach 2:
The defect sampling system performs self-service by automatically selecting defects without human intervention. The computer system uses stored design information and sampling algorithms to autonomously identify and select representative defects from array regions, making the process independent of manual operator input.
2Adaptability or versatility
If current manual sampling methods are used, then operators have control over defect selection, but the methods do not utilize available design information and array cell location data
Solution Approach 1:
The system performs preliminary action by pre-storing and organizing design information, including array cell type locations and specifications, before the sampling process. This pre-prepared information is then automatically retrieved and used during defect selection, eliminating the need for manual reference to design documents and ensuring consistent utilization of design data.
Solution Approach 2:
The sampling system achieves universality by integrating multiple functions: it stores design information, identifies array cell types, locates defects within those cells, and selects samples based on combined criteria. This multi-functional approach replaces multiple separate manual processes with a single integrated system that utilizes all available information.
3Reliability
If defects from all areas are included in the sample, then comprehensive defect coverage is achieved, but processing time increases and wasteful processing occurs when users only need defects from specific areas
Solution Approach 1:
The system applies local quality by allowing different sampling strategies for different areas or cell types within the array region. Users can specify which particular array cells or regions should be sampled, and the system adjusts its selection criteria accordingly, providing tailored sampling for specific areas rather than uniform sampling across all regions.
Solution Approach 2:
The system implements partial action by sampling only the necessary portion of defects based on user specifications. Instead of processing all detected defects, the system selectively samples only those from designated areas or cell types, reducing processing workload while maintaining adequate representation for the specific analysis needs.
Data Source
AI summary
Methods and systems for creating a sample of defects for a specimen are provided. One method includes detecting defects on a specimen based on output generated by a detector of an output acquisition subsystem. For the defects detected in an array region on the specimen, where the array region includes multiple array cell types, the method includes stacking information for the defects based on the multiple array cell types. The stacking includes overlaying design information for only a first of the multiple array cell types with the information for only the defects detected in the first of the multiple array cell types. In addition, the method includes selecting a portion of the detected defects based on results of the stacking thereby creating a sample of the detected defects.