Unveiling the layer-dependent electronic properties in transition-metal dichalcogenide heterostructures assisted by...

Area of Science:

• Materials Science
• Condensed Matter Physics
• Computational Chemistry

Background:

• Layer number (N) significantly influences electronic properties of 2D transition-metal dichalcogenide (TMD) van der Waals (vdW) heterostructures.
• Investigating layer-dependent properties across a wide range of N requires substantial computational resources.

Purpose of the Study:

• To develop a machine learning (ML) model for predicting layer-dependent electronic properties of TMD vdW heterostructures.
• To efficiently explore the impact of layer number (N) on electronic properties for MoS2, WS2, MoSe2, WSe2, MoTe2, and WTe2.

Main Methods:

• Density Functional Theory (DFT) calculations were combined with machine learning (ML) techniques.
• ML models were trained to predict bandgaps and band edge positions for heterostructures with N ranging from 2 to 10.

Main Results:

• ML models achieved cross-validation scores exceeding 90% for predicting bandgaps and band edge positions.
• Electronic properties are highly sensitive to layer number in few-layer systems, with significant changes observed between bilayer and triple-layer structures.
• TMD heterostructure properties stabilize for N > 8, with diminishing influence of layer number.

Conclusions:

• The study provides an efficient ML-driven approach to understand and predict layer-dependent electronic properties of multi-layer TMD vdW heterostructures.
• The findings accelerate the discovery of novel functional materials by reducing the need for extensive DFT calculations.