Wafer Normality Level Analysis Using Inline and Offline Measurements
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing defect inspection tools primarily focus on categorizing wafers as normal or abnormal without distinguishing between different levels of normality or abnormality, and do not provide a method to categorize wafers based on their quality or proximity to defects.
Innovation Solution
A system that utilizes machine learning to analyze the relationship between measurement data from inline and offline tools, estimating the normality or abnormality level of samples processed by processing tools, allowing for a more nuanced categorization of wafer quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If defect inspection tool focuses on extracting defective wafers and categorizes them as normal/abnormal, then defect detection efficiency is improved, but the ability to identify different levels of normality or abnormality is lost
Solution Approach 1:
The patent segments the traditional binary classification (normal/abnormal) into multiple granularity levels. It introduces intermediate categories such as 'normal with high risk', 'normal with low risk', and different abnormality levels, allowing the system to maintain efficient defect detection while preserving detailed normality level information for further analysis and differentiated handling.
Solution Approach 2:
The patent adds a new dimension to the classification system by introducing normality level assessment alongside the traditional defect detection. This multi-dimensional approach enables the system to evaluate both whether defects exist and the degree of normality, thereby resolving the contradiction between efficient defect detection and information preservation.
2Speed
If all normal wafers are uniformly categorized without further inspection, then processing speed is improved, but the ability to identify high-quality wafers or those close to defective is lost
Solution Approach 1:
The patent applies local quality assessment by evaluating different aspects of wafer normality separately. Instead of uniform categorization, it assesses specific characteristics and assigns different normality levels to different regions of the classification space, enabling precise identification of high-quality wafers while maintaining overall processing efficiency.
3Measurement precision
If machine learning is used to estimate normality level, then wafer quality differentiation is improved, but system complexity increases
Solution Approach 1:
The patent introduces a normality level estimation module as an intermediary between defect detection and final classification. This modular approach allows the system to incorporate sophisticated machine learning-based normality assessment without fundamentally redesigning the entire inspection system, thereby managing complexity while achieving precise normality level measurement.
Data Source
AI summary
An objective of the present disclosure is to provide a technique for identifying normality level/abnormality level of sample processed by processing tool. The system according to the present disclosure comprises: a learner configured to learn a relationship between measurement data that describes a measurement result acquired by a first measuring tool and test data that describes a measurement result acquired by a second measuring tool which measures the sample after a manufacturing process for the sample is finished; and a processor that estimates, using the learner, a normality level or an abnormality level of the sample measured by the second measuring tool.