Determination of the Interface between Amorphous Insulator and Crystalline 4H-SiC in Transmission Electron Microscope Image by using Convolutional Neural Network

TL;DR

This paper employs a convolutional neural network to accurately identify the amorphous-crystalline interface in TEM images of SiC, aiding in understanding low mobility issues in SiC MOSFETs.

Contribution

It introduces a CNN-based method for precise interface detection in TEM images, improving over manual boundary drawing especially for rough interfaces.

Findings

CNN accurately recognizes the interface in TEM images.

The method effectively handles rough interfaces that are difficult to delineate manually.

Power spectral density analysis of interface roughness was performed.

Abstract

A rough interface seems to be one of the possible reasons for low channel mobility (conductivity) in SiC MOSFETs. To evaluate the mobility by interface roughness, we drew a boundary line between amorphous insulator and crystalline 4H-SiC in a cross-sectional image obtained by a transmission electron microscope (TEM), by using the deep learning approach of convolutional neural network (CNN). We show that the CNN model recognizes the interface very well, even when the interface is too rough to draw the boundary line manually. Power spectral density of interface roughness was calculated.