Recipe Parameter Management for Semiconductor Defect Review
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing semiconductor manufacturing review systems face inefficiencies in determining optimal review conditions due to high operator load and reliance on trial reviews, particularly with Scanning Electron Microscopes (SEM) where contamination marks are common, leading to increased trial reviews and difficulty in defect image viewing.
Innovation Solution
A recipe parameter management system that collects and manages data on setting values, number of trial reviews, and defect images, generating histograms to display trends, allowing operators to easily access past parameter settings and recognize difficulty in parameter setting, thereby reducing the number of trial reviews needed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If trial reviews are conducted multiple times to determine optimal review conditions, then the review quality is improved, but the number of trial reviews increases and operator load increases
Solution Approach 1:
The system performs preliminary analysis by collecting historical recipe parameter data and defect image data before the actual review process. By pre-processing this data to identify contamination marks and determine optimal review conditions in advance, the system eliminates the need for multiple trial reviews during actual operation, thus improving review quality while reducing the number of trials needed
Solution Approach 2:
The system creates a virtual model of the review process by analyzing historical data and defect images to simulate optimal review conditions. This virtual copying of the review process allows operators to determine optimal parameters without conducting multiple physical trial reviews, reducing time loss while maintaining review quality
2Manufacturing precision
If trial reviews are conducted multiple times to determine optimal review conditions, then the review quality is improved, but operator load increases
Solution Approach 1:
The system performs self-service by automatically collecting recipe parameter data and defect image data, then autonomously analyzing this data to determine optimal review conditions. The system generates histograms and identifies contamination marks without operator intervention, thereby improving review quality while reducing operator load as the system serves itself
3Manufacturing precision
If SEM is used for defect review, then review capability is improved, but contamination marks appear on wafer surface
Solution Approach 1:
The system performs preliminary analysis of historical defect image data collected from SEM reviews to identify contamination marks before they affect subsequent reviews. By detecting and accounting for contamination patterns in advance through data analysis, the system maintains SEM review capability while compensating for contamination effects
Solution Approach 2:
The system establishes feedback by continuously collecting and analyzing defect image data from SEM reviews, using this information to identify contamination marks and adjust review parameters. This feedback loop allows the system to maintain high review capability while accounting for and compensating for contamination marks generated by SEM operation
Data Source
AI summary
In operation to obtain an optimal observation condition in a review system, the number of trial reviews can be reduced to improve efficiency of the operation. For a defect review conducted by the review system, a recipe parameter management system stores, as recipe parameter setting history in a recipe parameter setting history database (DB), a recipe parameter setting values of recipe parameters set when the defect review is conducted, the number of trial reviews carried out until the recipe parameter setting values are set, and defect images obtained when the defect review is conducted. The apparatus displays, on a terminal, histograms and the numbers of trial reviews generated based on the recipe parameter setting history data stored in the recipe parameter setting history database (DB). Hence, the operator can easily obtain data regarding the recipe parameter setting in the past.