AI-designed waveguides pave the way for next-generation photonic devices

A team of researchers at the University of California, Los Angeles (UCLA) has introduced a novel framework for designing and creating universal diffractive waveguides that can control the flow of light in highly specific and complex ways.

This new technology uses artificial intelligence (AI), specifically deep learning, to design a series of structured surfaces that guide light with high efficiency and can perform a wide range of functions that are challenging for conventional waveguides.

The work is published in the journal Nature Communications.

Optical waveguides, which are structures that guide light waves, are fundamental components in modern technology, essential for everything from global telecommunications networks to integrated photonic circuits and advanced sensors.

Traditional waveguides, like fiber optic cables, confine light within a core material that has a higher refractive index than its surrounding cladding, guiding light over long distances with minimal loss.

However, creating waveguides that can perform more complex tasks—such as filtering specific light modes, splitting them into different channels based on their polarization and/or spectrum, or bending light around sharp corners—often requires complex designs, specialized materials, and sophisticated fabrication processes.

The research, led by Professor Aydogan Ozcan of the UCLA Electrical and Computer Engineering Department, aims to overcome these challenges using a powerful, AI-driven design approach. Instead of relying on traditional materials to confine light, the new system uses a series of thin, transparent diffractive layers.

These cascadable layers, which can be thought of as smart, structured surfaces, are optimized by a deep learning algorithm to collectively sculpt and guide a light beam as it propagates. The AI fine-tunes the patterns on each surface to ensure that desired light modes pass through with minimal loss and high purity, while unwanted modes are filtered out.

“Our diffractive waveguide framework reimagines how we can control light. Instead of being constrained by the physical properties of materials, we can teach a sequence of surfaces to guide light and perform complex optical tasks in a cascaded manner,” explained Dr. Ozcan, the study’s corresponding author.

“This gives us a new toolbox, like an optical Lego set, to create highly versatile, task-specific waveguides that can be cascaded for a wide range of applications, from advanced optical communication systems to compact and sensitive sensors.”

The team demonstrated the power of their platform by designing several diffractive waveguides that perform specialized functions, including mode filters that selectively transmit or block specific spatial and spectral modes of light and mode-splitting waveguides that separate and multiplex different light modes into distinct output channels for communications.

Their designs also included waveguides for mode-specific polarization control, which maintain the desired polarization state of certain spatial modes while filtering out others.

One of the significant advantages of this technology is its scalability and versatility. A design that is optimized for one wavelength can be physically scaled to work in other parts of the electromagnetic spectrum, such as the visible or infrared, without needing to be redesigned or retrained. Furthermore, the diffractive waveguides can function in air or even when immersed in liquids or gases, opening up new applications in sensing.

This research was conducted by an interdisciplinary team from UCLA’s Electrical and Computer Engineering Department, Bioengineering Department, and the California NanoSystems Institute (CNSI) led by Professors Aydogan Ozcan and Mona Jarrahi of UCLA.