Vector vortex beams sorting of 120 modes in visible spectrum.

Area of Science:

• Optics and Photonics
• Artificial Intelligence in Optics

Background:

• Vector vortex beams possess three degrees of freedom: polarization (P), angular index (l), and radius index (p).
• Efficient sorting of individual degrees of freedom is established, but simultaneous sorting remains a significant challenge.
• Applications span optical manipulation, communication, and imaging.

Purpose of the Study:

• To propose and experimentally demonstrate a compact and efficient beam sorter capable of simultaneously handling all three degrees of freedom of vector vortex beams.
• To leverage diffractive deep neural networks (D²NNs) for advanced optical field manipulation.

Main Methods:

• Development of a novel beam sorter based on a diffractive deep neural network (D²NN).
• Experimental demonstration of the D²NN beam sorter in the visible spectrum.
• Utilizing the D²NN to sort vector vortex beams based on polarization, angular, and radius indices.

Main Results:

• Successful and robust simultaneous sorting of 120 distinct Laguerre-Gaussian (LG) modes.
• Demonstration of the D²NN's capability for complex optical field sorting.
• Validation of the proposed method in the visible light spectrum.

Conclusions:

• The proposed D²NN-based beam sorter offers a compact and efficient solution for simultaneous sorting of multiple degrees of freedom in vector vortex beams.
• Diffractive deep neural networks show significant potential for advanced optical field manipulation.
• This technology can enhance diverse applications of vector vortex beams.