Machine Learning-Assisted Light Management and Electromagnetic Field Modulation of Large-Area Plasmonic Coaxial...

Area of Science:

• Plasmonics and Nanophotonics
• Computational Electromagnetics
• Materials Science

Background:

• Fabricating hexagonal coaxial cylindrical gold pillar/ring nanoarrays is challenging experimentally.
• Optimizing geometrical parameters for desired optical properties requires extensive, costly experimentation.

Purpose of the Study:

• Investigate the influence of geometrical parameters on optical properties of gold nanoarrays.
• Utilize simulations and machine learning to guide the design of nanoarrays for specific optical applications.

Main Methods:

• Employed finite-difference time-domain (FDTD) simulations to model optical responses.
• Analyzed localized surface plasmon resonance (LSPR) and charge distributions.
• Applied machine learning for parameter analysis and visualization.

Main Results:

• Optical properties like plasmonic resonance band, electric field enhancement, and Q-factor are tunable.
• Radiative damping can be suppressed, and electric fields concentrated by managing LSPR and charge distribution.
• Nanoarray height and gap width are key parameters for optimizing Q-factor and electric field enhancement.

Conclusions:

• FDTD simulations coupled with machine learning offer a theoretical framework for designing nanoarrays.
• This approach facilitates tailoring geometrical parameters for specific optical functionalities.
• Provides insights into optimizing plasmonic nanostructures for enhanced optical performance.