Search research articles

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

Neuroevolution-Based Network Architecture Evolution in Semiconductor Manufacturing.

ACS omega·2023

Same author

Meta-Learned and TCAD-Assisted Sampling in Semiconductor Laser Annealing.

ACS omega·2023

Same author

An ultra-compact blackbody using electrophoretic deposited carbon nanotube films.

RSC advances·2022

Same author

Analytics-statistics mixed training and its fitness to semisupervised manufacturing.

PloS one·2019

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Video

Updated: Oct 6, 2025

Silicon Metal-oxide-semiconductor Quantum Dots for Single-electron Pumping

Silicon Metal-oxide-semiconductor Quantum Dots for Single-electron Pumping

Published on: June 3, 2015

Mapping Oxidation and Wafer Cleaning to Device Characteristics Using Physics-Assisted Machine Learning.

Sparsh Pratik¹, Po-Ning Liu¹, Jun Ota^1,2

¹Institute of Electronics Engineering, National Yang Ming Chiao-Tung University, Hsinchu City 30010, Taiwan, R.O.C.

|January 17, 2022

Summary

This summary is machine-generated.

This study introduces a physics-assisted machine learning approach to link semiconductor processing conditions with device capacitance-voltage (C-V) characteristics. The novel method significantly improves prediction accuracy compared to traditional models.

More Related Videos

The Effect of Anodization Parameters on the Aluminum Oxide Dielectric Layer of Thin-Film Transistors

The Effect of Anodization Parameters on the Aluminum Oxide Dielectric Layer of Thin-Film Transistors

Published on: May 24, 2020

Environmental Dynamic Mechanical Analysis to Predict the Softening Behavior of Neural Implants

Environmental Dynamic Mechanical Analysis to Predict the Softening Behavior of Neural Implants

Published on: March 1, 2019

Related Experiment Videos

Last Updated: Oct 6, 2025

Silicon Metal-oxide-semiconductor Quantum Dots for Single-electron Pumping

Silicon Metal-oxide-semiconductor Quantum Dots for Single-electron Pumping

Published on: June 3, 2015

The Effect of Anodization Parameters on the Aluminum Oxide Dielectric Layer of Thin-Film Transistors

The Effect of Anodization Parameters on the Aluminum Oxide Dielectric Layer of Thin-Film Transistors

Published on: May 24, 2020

Environmental Dynamic Mechanical Analysis to Predict the Softening Behavior of Neural Implants

Environmental Dynamic Mechanical Analysis to Predict the Softening Behavior of Neural Implants

Published on: March 1, 2019

Area of Science:

Materials Science
Electrical Engineering
Artificial Intelligence

Background:

Establishing a clear link between semiconductor processing chemistry and device performance is crucial, especially with shrinking technology nodes.
Traditional physical models struggle to capture the complex relationships between process parameters (e.g., oxidation, cleaning) and device electrical characteristics.
Machine learning (ML) and hybrid approaches are gaining traction but require substantial data, which is challenging to generate in semiconductor manufacturing.

Purpose of the Study:

To develop and evaluate a novel physics-assisted machine learning approach for modeling the relationship between semiconductor process conditions and device capacitance-voltage (C-V) curves.
To address data sparsity issues in semiconductor manufacturing by leveraging physics-assisted artificial intelligence.

Main Methods:

Utilized a physics-assisted multitask and transfer learning approach.
Developed a machine learning architecture capable of handling data sparsity.
Compared the performance of semisupervised multitask learning (SS-MTL) and transfer learning (TL) against a pure multilayer perceptron (MLP) model.

Main Results:

The physics-assisted ML models demonstrated significant improvements in predicting C-V curves.
Achieved a coefficient of determination (R²) of 0.9442 for SS-MTL and 0.9253 for TL.
Outperformed the pure ML model, which had an R² of 0.6108.

Conclusions:

Physics-assisted ML effectively models the complex relationship between semiconductor manufacturing processes and device characteristics, even with limited data.
The developed ML architecture successfully handles data sparsity and enhances predictive accuracy for C-V curves.
This hybrid approach offers a promising solution for semiconductor process and device modeling.