Unraveling the Correlation between Raman and Photoluminescence in Monolayer MoS₂ through Machine-Learning Models.

Area of Science:

• Materials Science
• Condensed Matter Physics
• Nanotechnology

Background:

• Two-dimensional transition metal dichalcogenides (TMDCs) exhibit tunable photoluminescence (PL), enabling advanced optoelectronic and photonic devices.
• Raman spectroscopy is crucial for characterizing 2D materials' crystallinity, doping, and strain.
• The nonlinear correlation between PL and Raman spectra in monolayer MoS2 requires deeper investigation.

Purpose of the Study:

• To systematically explore the connections between photoluminescence (PL) signatures and Raman modes in monolayer MoS2.
• To provide comprehensive insights into the physical mechanisms linking PL and Raman features.
• To disentangle strain and doping contributions within Raman spectra using machine learning.

Main Methods:

• Utilized a dense convolutional network (DenseNet) for predicting PL maps from spatial Raman maps.
• Employed a gradient boosted trees model (XGBoost) with Shapley additive explanation (SHAP) to analyze Raman feature impacts on PL.
• Applied a support vector machine (SVM) to project PL features onto Raman frequencies.

Main Results:

• Established a systematic methodology for correlating PL and Raman spectra in monolayer MoS2.
• Successfully disentangled the effects of strain and doping on Raman spectra via machine learning models.
• Demonstrated the predictive power of Raman spectroscopy for understanding PL characteristics.

Conclusions:

• The study provides a robust framework for analyzing the interplay between PL and Raman spectra in 2D materials.
• Machine learning offers a powerful approach for advanced characterization and understanding of TMDCs.
• This methodology can be extended to other 2D materials for material property investigation.