Researchers developed a new type of integrated photonic device capable of generating broad ranges of light frequencies on a chip. The device uses a silicon nitride core to generate optical frequency combs while a surrounding silica layer produces Raman scattering. Credit: Alekhya Ghosh

A layer once dismissed as mere support has transformed a tiny photonic chip into a powerful generator of new light frequencies.

A laser usually produces one narrow color of light. A device small enough to sit on a fingertip can turn that single input into hundreds of precisely spaced frequencies, creating a tool for measuring time, identifying chemicals, transmitting data, and studying distant objects in space.

Researchers have now expanded what such photonic chips can do by making two of their materials work together. Instead of treating the outer layer as simple packaging, the team used it to generate additional light frequencies that the chip’s main material could not efficiently produce on its own.

The device combines a silicon nitride core with a surrounding layer of silica. Inside the core, the light undergoes a nonlinear process that can produce an optical frequency comb, a series of evenly spaced frequencies often compared to the markings on a ruler. In the silica, a second process called Raman scattering shifts some of the light to new frequencies.

Bringing these effects together allowed the researchers to demonstrate Raman lasing in a silica-clad silicon nitride photonic circuit and then use the interaction between the two processes to generate broadband frequency combs. The findings were reported in Advanced Photonics.

Making a Supposedly Passive Layer Useful

Photonic chips guide light through microscopic channels called waveguides. Most of the optical energy remains in the waveguide core, while a cladding layer around it helps keep the light confined.

A portion of the optical field extends beyond the core and enters the cladding. Engineers often treat this faint spillover as a side effect, but the researchers recognized it as an opportunity.

They designed a ring-shaped silicon nitride resonator in which about 31 percent of the circulating optical field overlaps with the surrounding silica. As the light repeatedly travels around the ring, it interacts with both materials rather than only with the core.

Two Materials, Two Complementary Roles

Silicon nitride provides the mechanism for generating an optical frequency comb. Its strong Kerr nonlinearity allows light from a continuous wave laser to interact through processes such as four-wave mixing, producing many evenly spaced frequencies around the original laser signal.

Silica performs a different function. It supplies the Raman response needed to create light at a separate, shifted frequency. During Raman scattering, incoming light exchanges energy with molecular vibrations in the silica. If that shifted light is amplified strongly enough, the resulting Raman gain can produce Raman lasing.

Instead of forcing one material to perform both jobs, the chip lets each one handle the effect it produces best.

A New Signal Emerges

To test whether combining silicon nitride and silica could produce Raman lasing, the researchers sent continuous-wave laser light into silica-coated silicon nitride ring resonators. A second optical signal appeared 11 terahertz away from the pump frequency, matching the characteristic Raman shift of silica.

When the researchers changed the pump wavelength, the new signal moved with it while maintaining the same 11 terahertz separation. This consistent offset confirmed that the added light came from Raman scattering in the silica rather than from an unrelated effect.

At first, Raman lasing produced Stokes and anti-Stokes sidebands. These signals appear at frequencies below and above the original laser frequency. At higher power, four-wave mixing grew stronger in the silicon nitride core, producing additional frequencies around both the pump light and the Raman-shifted signals.

Further interactions caused those frequencies to multiply, ultimately forming broad clusters of evenly spaced comb lines.

Waveguide Geometry Unlocks a Wider Spectrum

The team widened the silicon nitride waveguide slightly to control dispersion, which describes how different wavelengths travel through a material. This small geometric change improved the alignment between optical modes and allowed the nonlinear interactions to occur more efficiently.

With the modified resonator, the frequency comb extended across more than 400 nanometers. Comb lines formed not only near the original pump wavelength but also around several Raman-shifted frequencies, greatly expanding the spectral output.

Different wavelengths can serve different purposes. In spectroscopy, for example, molecules absorb distinct frequencies that act like chemical fingerprints. A broadband comb can probe many of those signatures at once. In communications, individual comb lines can potentially function as separate data channels.

Optical frequency combs are also important in precision measurement, including optical clocks, astronomical instruments, and systems that compare widely separated frequencies. Conventional comb generators can occupy a laboratory bench. Microresonators offer a path toward much smaller and potentially more practical versions.

Results Closely Match Theory

The researchers calculated that Raman lasing should begin when the optical power reaching the chip approached 140 milliwatts. In the experiment, the measured threshold was 143 milliwatts.

That close agreement provided strong evidence that the silica cladding was responsible for the Raman gain. The researchers also reported that improving the resonator’s optical quality could eventually lower the required power to the milliwatt range.

The generated combs were not fully coherent, meaning the phases of all the comb lines were not perfectly locked together. Full coherence is important for the most demanding timing and measurement applications, so further engineering will be needed.

Even so, the system converted more than 32 percent of the incoming optical power into newly generated frequencies. That efficiency suggests the hybrid approach can produce substantial output rather than only a weak laboratory signal.

A Broader Strategy for Photonic Chips

Integrated photonics has often focused on finding one material that combines low loss, easy manufacturing, wide transparency, and strong nonlinear behavior. No material excels at every task. Adding separate components can solve the problem, but it can also make a chip larger, harder to manufacture, or more difficult to operate.

The new work points toward another option: use the existing layers of a photonic circuit as a coordinated system.

Similar designs could combine other materials with complementary optical properties, potentially supporting broadband supercontinuum sources, tunable lasers, or self-referenced frequency combs. A self-referenced comb can determine its own absolute frequency positions, a capability needed for highly precise clocks and measurements.

Reference: “Hybrid nonlinear effects in photonic integrated circuits” by Arghadeep Pal, Alekhya Ghosh, Shuangyou Zhang, Toby Bi, Masoud Kheyri, Haochen Yan, Yaojing Zhang and Pascal Del’Haye, 23 June 2026, Advanced Photonics.
DOI: 10.1117/1.AP.8.4.046008

Never miss a breakthrough: Join the SciTechDaily newsletter.
Follow us on Google and Google News.