Dispersion engineered metasurfaces for broadband, high-NA, high-efficiency, dual-polarization analog image processing

Area of Science:

• Photonics and Optical Engineering
• Metasurface Technology
• Analog Image Processing

Background:

• Optical metasurfaces offer potential for compact analog image processing, reducing power and latency.
• Existing metasurface designs face limitations in resolution, throughput, polarization, bandwidth, and isotropy.
• Addressing these trade-offs is crucial for practical optical computing applications.

Purpose of the Study:

• To demonstrate a novel approach for optimizing metasurface performance metrics simultaneously.
• To engineer silicon metasurfaces capable of high-performance analog image processing.
• To introduce quantitative efficiency metrics for evaluating metasurface devices.

Main Methods:

• Utilizing dispersion engineering to design and optimize optical metasurfaces.
• Experimentally fabricating and characterizing silicon-based metasurface devices.
• Implementing quantitative metrics to assess device efficiency and performance.

Main Results:

• Achieved simultaneous optimization of spatial resolution, throughput, polarization, bandwidth, and isotropy.
• Demonstrated silicon metasurfaces for isotropic and dual-polarization edge detection.
• Exhibited numerical apertures > 0.35 and spectral bandwidths of 35 nm around 1500 nm.
• Reported high throughput efficiencies, approaching theoretical limits for given numerical apertures.

Conclusions:

• Dispersion engineering provides an effective strategy for overcoming metasurface design trade-offs.
• The developed silicon metasurfaces enable efficient, broadband, and polarization-insensitive optical image processing.
• These advancements facilitate the development of low-loss, high-efficiency optical computing and free-space image processing systems.