# The Impact of Microwave Annealing on MoS2 Devices Assisted by Neural Network-Based Big Data Analysis

## TL;DR

This paper explores how microwave annealing affects MoS2 devices using a neural network to improve uniformity and determine optimal annealing power.

## Contribution

A novel neural network-based method using HSV color space is introduced to enhance MoS2 device uniformity and determine optimal microwave annealing power.

## Key findings

- A neural network using HSV color space effectively distinguishes MoS2 film thickness for uniform device fabrication.
- Optimal microwave annealing power for MoS2 devices was identified as 700 W based on electrical performance analysis.

## Abstract

Microwave annealing, an emerging annealing method known for its efficiency and low thermal budget, has established a foundational research base in the annealing of molybdenum disulfide (MoS2) devices. Typically, to obtain high-quality MoS2 devices, mechanical exfoliation is commonly employed. This method’s challenge lies in achieving uniform film thickness, which limits the use of extensive data for studying the effects of microwave annealing on the MoS2 devices. In this experiment, we utilized a neural network approach based on the HSV (hue, saturation, value) color space to assist in distinguishing film thickness for the fabrication of numerous MoS2 devices with enhanced uniformity and consistency. This method allowed us to precisely assess the impact of microwave annealing on device performance. We discovered a relationship between the device’s electrical performance and the annealing power. By analyzing the statistical data of these electrical parameters, we identified the optimal annealing power for MoS2 devices as 700 W, providing insights and guidance for the microwave annealing process of two-dimensional materials.