Pulsed lasers have been the workhorse of ultrafast optics since their advent and rapid development throughout the 20th century, revolutionizing a wide variety of fields from spectroscopy to tattoo removal. Over the past twenty years, pulsed laser sources have shrunk from tabletop laboratory setups down to micron-sized chips, making them ideal components for integrated photonic devices. Despite this miniaturization, chip-scale pulsed laser sources have eluded the mid infrared (IR) spectral region. Active mode-locking of mid IR semiconductor lasers—such as quantum cascade lasers (QCLs)—has produced pulse widths on the order of 6 picoseconds [1]. Pulse compression techniques can be utilized to shrink these pulses to hundreds of femtoseconds [2], but rely on large optical setups that cannot be scaled down. Here, we present a fundamentally new way to produce bright pulses of mid IR light by optically pumping ring QCLs. This technique unifies the physics of passive, Kerr microresonator combs and ring QCLs [3]. Using a modified racetrack QCL with an integrated directional coupler, we injection-lock the unidirectional laser field circulating in the racetrack to a commercial external cavity QCL. Much like in Kerr microresonator combs, when the injection-locked field is detuned from its natural cavity resonance, the resonance becomes bistable, with its unstable branch supporting bright solitons with pulse widths of ~1 picosecond at a center wavelength of 8 µm. This method of pulse formation is well-suited for lasers with fast gain dynamics, which encompasses the entire family of QCLs, spanning from 3 µm to 300 µm. Furthermore, the optical drive can, in principle, be integrated with the racetrack, providing a route for on-chip, ultrashort pulse formation throughout the entire mid-IR.

[1] J. Hillbrand, et. al., Nat. Commun. 11, 5788(2020).

[2] P. Täschler, et. al., Nat. Photon. 15, 919-924(2021).

[3] L. Columbo, M. Piccardo, et. al., Phys. Rev. Lett. 126, 173903(2021).

Compact and efficient mid-infrared (MIR) frequency combs are expected to find widespread use in chemical sensing applications, such as on-chip spectroscopy of toxic substances. While most of the experimental MIR laser comb work has involved quantum cascade lasers (QCLs), interband cascade lasers (ICLs) operate cw at room temperature in the 3-4 mm spectral range, which remains difficult for QCLs, and also promise significant reductions in the operating power throughout the MIR spectral range. Furthermore, owing to the long carrier lifetime in an ICL, both passive mode-locking with short pulses (AM) and quasi-cw (FM) comb generation should be possible.

In spite of these promising characteristics, only FM combs based on ICLs have been demonstrated to date. In order to clarify the physical requirements for both AM and FM operation, we have developed a multiscale numerical model that efficiently solves the Maxwell-Bloch equations for the full subband dispersion in the ICL’s active type-II wells over a time period of μs. We compare the results of this model to those derived from the two-level approximation relevant to QCLs, and evaluate the importance of such parameters as the second-order and higher-order group velocity dispersions, saturable absorber length and recovery time, ambipolar diffusion coefficient, and polarization relaxation time (homogeneous gain broadening linewidth). We determine the optimal design parameters for experimentally demonstrating both passively mode-locked and FM ICL combs, and outline how they can be realized in practice.