Optical circular dichroism engineering in chiral metamaterials utilizing a deep learning network.

Area of Science:

• Optics and Photonics
• Materials Science
• Computational Physics

Background:

• Chiral metamaterials exhibit unique light-matter interactions.
• Predicting chiroptical responses is crucial for designing advanced photonic devices.
• Traditional methods like RCWA are computationally intensive.

Purpose of the Study:

• To develop and validate a deep learning (DL) algorithm for predicting the chiroptical response of 2D chiral metamaterials.
• To compare the DL approach with the rigorous coupled wave analysis (RCWA) method.
• To demonstrate the potential of DL in accelerating metamaterial design.

Main Methods:

• A deep neural network (DL) model was developed to predict chiroptical responses.
• Rigorous coupled wave analysis (RCWA) was used for comparison.
• The study focused on 2D chiral metamaterials with U-like, T-like, and I-like nanostructures.
• Circular dichroism (CD) in higher-order diffracted beams was analyzed.

Main Results:

• The DL model accurately predicted the chiroptical response (CD) of 2D chiral metamaterials.
• A common feature observed was the strongest CD response in third-order diffracted beams.
• The DL approach achieved a computational speed four orders of magnitude faster than RCWA.

Conclusions:

• Deep learning offers a highly accurate and significantly faster alternative to RCWA for predicting chiroptical responses.
• The developed DL model shows great potential for exploring chiroptical interactions in metamaterials.
• This work accelerates the design of hypersensitive photonic devices utilizing chiral metamaterials.