Virtual Metrology models for predicting physical measurement in semiconductor manufacturing

The semiconductor manufacturing industry has a large-volume multistage manufacturing system. To insure the high stability and the production yield on-line a reliable wafer monitoring is required. The approach, called Virtual Metrology (VM) is defined as the prediction of metrology variables (either...

Full description

Saved in:
Bibliographic Details
Main Authors: Ferreira, A., Roussy, A., Conde, L.
Format: Conference Proceeding
Language:English
Subjects:
Current measurement
Manufacturing industries
Manufacturing systems
Metrology
Predictive models
Production
Semiconductor device manufacture
Semiconductor device modeling
Stability
Virtual manufacturing
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The semiconductor manufacturing industry has a large-volume multistage manufacturing system. To insure the high stability and the production yield on-line a reliable wafer monitoring is required. The approach, called Virtual Metrology (VM) is defined as the prediction of metrology variables (either measurable or non measurable) using process and wafer state information. It consists in the definition and the application of some predictive and corrective models for metrology outputs (physical measurements) in function of the previous metrology outputs and of the equipment parameters of current and previous steps of fabrication. The goals of this paper are to present a methodology for VM module for individual process applications in semiconductor manufacturing and to present a case study based on industrial data.
ISSN:1078-8743
2376-6697
DOI:10.1109/ASMC.2009.5155973