Deep learning-based quality filtering of mechanically exfoliated 2D crystals

TL;DR

This paper introduces a deep learning method using U-Net to automatically identify and segment monolayer and bilayer 2D crystals in optical images, significantly reducing manual effort in material screening.

Contribution

The study presents a novel deep learning approach that accurately classifies 2D crystal thicknesses from optical images, enabling high-throughput screening in material research.

Findings

Achieved 70% success rate in distinguishing monolayer and bilayer MoS2.

Demonstrated that AI can replace manual screening in 2D material fabrication.

Used only 24 images for training, showing efficiency of the method.

Abstract

Two-dimensional (2D) crystals are attracting growing interest in various research fields such as engineering, physics, chemistry, pharmacy and biology owing to their low dimensionality and dramatic change of properties compared to the bulk counterparts. Among the various techniques used to manufacture 2D crystals, mechanical exfoliation has been essential to practical applications and fundamental research. However, mechanically exfoliated crystals on substrates contain relatively thick flakes that must be found and removed manually, limiting high-throughput manufacturing of atomic 2D crystals and van der Waals heterostructures. Here we present a deep learning-based method to segment and identify the thickness of atomic layer flakes from optical microscopy images. Through carefully designing a neural network based on U-Net, we found that our neural network based on U-net trained only with the data based on 24 images successfully distinguish monolayer and bilayer MoS2 with a success rate of 70%, which is a practical value in the first screening process for choosing monolayer and bilayer flakes of MoS2 of all flakes on substrates without human eye. The remarkable results highlight the possibility that a large fraction of manual laboratory work can be replaced by AI-based systems, boosting productivity.