Optical frequency combs consist of light frequencies made of equidistant laser lines. The discovery of ''soliton microcombs'' by Professor Tobias Kippenberg's lab at EPFL in the past decade has enabled frequency combs to be generated on chip. In this scheme, a single-frequency laser is converted into ultra-short pulses called dissipative Kerr solitons.
Soliton microcombs are chip-scale frequency combs that are compact, consume low power, and exhibit broad bandwidth. Combined with large spacing of comb "teeth", microcombs are uniquely suited for a wide variety of applications, such as terabit-per-second coherent communication in data centers, astronomical spectrometer calibration for exoplanet searches and neuromorphic computing, optical atomic clocks, absolute frequency synthesis, and parallel coherent LiDAR.
However, one outstanding challenge is the integration of laser sources. While microcombs are generated on-chip via parametric frequency conversion (two photons of one frequency are annihilated, and a pair of two new photons are generated at a higher and lower frequency), the pump lasers are typically off-chip and bulky. Integrating microcombs and lasers on the same chip can enable high-volume production of soliton microcombs using well-established CMOS techniques developed for silicon photonics, however this has been an outstanding challenge for the past decade.
For the nonlinear optical microresonators, where soliton microcombs are formed, silicon nitride (Si3N4) has emerged as the leading platform due to its ultralow loss, wide transparency window from visible to mid-infrared, absence of two-photon absorption, and high power-handling capability. But achieving ultralow-loss Si3N4 microresonators is still insufficient for high-volume production of chip-scale soliton microcombs, as co-integration of chip-scale driving lasers are required.