In the past few years, there have been some encouraging developments on the road toward integrated circuits that operate on photons rather than electrons—so-called photonic ICs. Most notable among these have been developments in plasmonics, which have made it possible to shrink the wavelengths of light to fit into the tight dimensions of today’s ICs.
While graphene and silicon together have been partly responsible for enabling these advancements in plasmonics, silicon has not been so friendly to the lasers that would provide the light source for these photonic ICs. One of silicon’s intrinsic electronic properties, its indirect band gap, makes its ability to generate light from an electrical current pretty feeble.
Now, researchers at Yale University may have found a solution to silicon’s poor laser performance. They’ve turned to an unusual type of laser that uses not just light but sound to create the optical amplification: a Brillouin laser.
Brillouin lasers—like all other lasers—have two essential working parts: an optical cavity, which is an arrangement of mirrors; and a gain medium—say, crystals, gas, or a semiconductor—that produces optical amplification. When the optical gain is sufficient to compensate for the round-trip losses experienced by the light field, these two elements produce a positive feedback loop. The result: optical self-oscillation and laser emission.
The Brillouin laser is a bit different. It exploits a phenomenon known as Brillouin scattering, in which sound waves change the refractive index in a material. This combination of laser light and sound create the amplification required for lasing.
Sounds great. Light and sound joined together to amplify the light and achieve lasing. However, there’s been a big problem: Brillouin scattering is extremely weak in conventional silicon photonic waveguides, a hindrance which has stagnated development of silicon-based Brillouin lasers.
The Yale researchers described in the journal Science how they sidestepped this issue.
“To achieve strong Brillouin amplification in silicon, we have developed a special type of suspended nanoscale waveguide—termed a hybrid photonic-phononic waveguide—that is engineered to simultaneously guide light and sound waves, says Nils Otterstrom, a PhD student at Yale and co-author of the paper based on the research. “This MEMS-like optomechanical structure has allowed us to transform Brillouin interactions, once entirely absent from silicon photonic circuits, into the strongest and most tailorable gain mechanism in silicon.”
In addition to the challenge of developing the low-loss waveguide, the Yale researchers had to devise a new laser design that is able to harness a new type of forward Brillouin interaction. Traditional Brillouin lasers yield laser emission in the backward direction. With this new amplification process, laser emission co-propagates with the pump wave. This eliminates the need for on-chip isolators or circulators that are often required for traditional Brillouin lasers based on backward Brillouin scattering, according to Otterstrom.
The silicon Brillouin laser Otterstrom and his colleagues used in their demonstrations has been integrated with photonic mode filters and grating couplers. Because this device was fabricated using a standard silicon-on-insulator wafer, it is straightforward to combine these silicon-based Brillouin laser technologies with other photonic components—such as modulators and mode multiplexers—in the same active silicon layer, says Otterstrom.
However, he cautions that for mass-scale production with CMOS circuitry on the same chip, it will become necessary to employ a hybrid CMOS-MEMS process. Also, Brillouin lasers will never operate on their own since they require optical pumping to give enhanced performance for desired applications such as gyroscopes. “Brillouin lasers are complementary laser technologies that can augment on-chip capabilities,” said Otterstrom.
Looking forward, Otterstrom envisions a number of opportunities for silicon-based Brillouin lasers for a range of on-chip applications.
Otterstrom added: “These same device designs lend themselves to new types of on-chip amplifiers and filters that yield enhanced performance in the context of optical signal processing, metrology, and microwave photonics.”
For example, an important goal involves shaping the dynamics of this laser so that it acts as a ‘noise-eater.’
Dexter Johnson is a contributing editor at IEEE Spectrum, with a focus on nanotechnology.