Machine Learning based CVD Virtual Metrology in Mass Produced Semiconductor Process

TL;DR

This paper evaluates machine learning techniques for CVD virtual metrology, demonstrating that nonlinear methods with data imputation significantly improve prediction accuracy and reduce process variation in semiconductor manufacturing.

Contribution

It introduces a combined nonlinear feature selection and regression approach with data imputation for enhanced CVD VM accuracy.

Findings

Nonlinear algorithms outperform linear methods in VM accuracy.

Data imputation improves prediction when data availability is limited.

Achieved up to 0.7 prediction accuracy, reducing process variation by 70%.

Abstract

A cross-benchmark has been done on three critical aspects, data imputing, feature selection and regression algorithms, for machine learning based chemical vapor deposition (CVD) virtual metrology (VM). The result reveals that linear feature selection regression algorithm would extensively under-fit the VM data. Data imputing is also necessary to achieve a higher prediction accuracy as the data availability is only ~70% when optimal accuracy is obtained. This work suggests a nonlinear feature selection and regression algorithm combined with nearest data imputing algorithm would provide a prediction accuracy as high as 0.7. This would lead to 70% reduced CVD processing variation, which is believed to will lead to reduced frequency of physical metrology as well as more reliable mass-produced wafer with improved quality.