Coupled optical resonators can have interesting properties, including transmission spectra that imitate the eigenfrequencies produced by molecules. Researchers from Chalmers University of Technology (Göteborg, Sweden) have now coupled two microresonators to create a frequency comb on a chip (microcomb) that could enable some advanced applications.1
A frequency comb is an alternative to mode-locked lasers that can generate repetitive pulses of light at high rates; the Fourier optical properties of such a pulse train dictate that a myriad of evenly spaced optical frequencies is produced. Such a comb can be used to measure or generate frequencies with extreme precision. A microcomb is specifically generated by sending laser light to a microresonator optical cavity. Microcombs have two important attributes that make them extremely attractive for practical purposes: the frequency spacing between markers is very large (typically between 10 and 1000 GHz), much higher than the spacing in mode-locked laser frequency combs, and they can be implemented in photonic integrated circuits.
The compatibility with photonic integration brings benefits in terms of reduction of size, power consumption, and the possibility to reach mass-market applications. The large spacing between the comb teeth means that microcombs can be used for applications such as light sources for wavelength-division-multiplexed (WDM) fiber-optic communication systems or for the synthesis of pure microwave electromagnetic radiation.
The key to the new enhanced microcomb from Chalmers is the fact that the researchers have coupled two microresonators together. The microresonators interact with each other, similar to how atoms bind together when forming a diatomic molecule. The arrangement is known as a photonic molecule and has unique physical characteristics. The new microcomb is a coherent, tunable, and reproducible device with up to 10X higher net conversion efficiency than the current state of the art.
Silicon nitride construction
Fabricated in silicon nitride, two dispersion-engineered microring resonators are placed side by side; one microring has an input waveguide placed close to it and coupled via evanescence, while the other contains a heater, which controls the refractive index and thus the frequency location of the degeneracy point between the two resonators. Pumped with a mode-locked fiber laser and interfered with a molecular-absorption line near 1550.5156 nm for calibration, the photonic-molecule microcomb is tunable from 1520 to 1570 nm (center wavelength) and produces a spectrally flat frequency comb that spans about 100 nm around the center wavelength.
The intrinsic quality factor (Q) was 7.5 × 106 and 5.7 × 106 for two different devices, while the extrinsic Q was 3.8 × 106 and 3.1 × 106 for the same devices. An experimental plot of a resonance and a counterpropagating mode could be fitted with a curve modeling that of a resonance doublet (see figure). The pump laser power required was only about 13 mW.
“The reason why the results are important is that they represent a unique combination of characteristics in terms of efficiency, low-power operation, and control that are unprecedented in the field,” says Óskar Bjarki Helgason, a PhD student at the Department of Microtechnology and Nanoscience at Chalmers.
These microcombs could, for example, radically decrease the power consumption in optical communications systems, with tens of lasers being replaced by a single chip-scale microcomb in datacenter interconnects. They could also be used in lidar for autonomous driving vehicles. Another area where microcombs could be used is for the calibration of the spectrographs used in astronomical observatories devoted to the discovery of Earth-like exoplanets. Extremely accurate optical clocks and spectroscopic health-monitoring apps for mobile phones are further possibilities.
“For the technology to be practical and find its use outside the lab, we need to co-integrate additional elements with the microresonators, such as lasers, modulators, and control electronics,” says Victor Torres Company, who leads the research project at Chalmers. “This is a huge challenge, that requires maybe five to 10 years and an investment in engineering research. But I am convinced that it will happen. The most interesting advances and applications are the ones that we have not even conceived of yet. This will likely be enabled by the possibility of having multiple microcombs on the same chip.”
REFERENCE
1. Ó. B. Helgason et al., Nat. Photonics (2021); https://doi.org/10.1038/s41566-020-00757-9.
