Visualized and Nondestructive Quality Identification of Two-Dimensional MoS₂ Based on Principal Component Analysis

Area of Science:

• Materials Science
• Nanotechnology
• Machine Learning Applications

Background:

• Current quality assessments for molybdenum disulfide (MoS2) are often destructive, inefficient, and lack scalability.
• There is a critical need for non-destructive, high-throughput methods to characterize MoS2 quality.

Purpose of the Study:

• To develop a visualized and nondestructive approach for evaluating MoS2 quality using machine learning.
• To differentiate MoS2 samples based on varying edge defect densities and identify structural features like grain boundaries.

Main Methods:

• Principal Component Analysis (PCA) applied to photoluminescence (PL) mapping data.
• Non-destructive characterization techniques including lifetime mapping and thermal expansion coefficient measurements.

Main Results:

• PCA effectively distinguished CVD-grown MoS2 with different edge defect densities.
• Six twin grain boundaries in MoS2 stars were successfully identified.
• High-quality MoS2 samples exhibited shorter carrier lifetimes (~0.291 ns) and lower thermal expansion coefficients (~2.03 × 10⁻⁵ K⁻¹).

Conclusions:

• The proposed PCA-based method provides an efficient and non-destructive way to evaluate MoS2 quality.
• This approach enables better material selection and quality control for MoS2 applications.
• The findings support the use of machine learning for advanced materials characterization.