Magnetically tunable random lasing in polymer-dispersed chiral liquid crystals: Experimental and machine learning insights

Abstract

This work investigates magnetically tunable random lasing in polymer-dispersed chiral liquid crystals (PDCLCs) doped with ferromagnetic nanoparticles (MNPs) in a capillary system. The alignment of liquid crystal molecules, modulated by external magnetic fields, strongly influences the lasing dynamics. We systematically studied the effects of MNP concentration and the amplitude and orientation of the magnetic field on key emission characteristics, including the lasing threshold, the spectral envelope, and intensity. To model these nonlinear interactions and permit prediction, supervised regression models were trained using experimental data. Four machine-learning algorithms (random forest, support vector machines, artificial neural networks, and XGBoost) were tested, and the XGBoost model yielded the best results (R2=0.9752) on the test set. Importance analysis revealed that magnetic-field strength and MNP concentration are the dominant factors governing lasing behavior. The results highlight the intricate interplay between magnetic-field modulation and nanoparticle doping in PDCLC-based random lasers, underscoring the potential of machine learning in optimizing future optoelectronic and photonic devices.