Terahertz (THz) quantum-cascade lasers (QCLs) are an emerging semiconductor source of compact, high-power THz radiation. Though first realized more than 15 years ago, THz QCLs continue to suffer from poor beam quality and outcoupling efficiency due to the subwavelength nature of the semiconductor ridge-waveguides typically used. In this thesis, a new technique is discussed for obtaining high power and good beam quality from THz QCLs, the THz quantum-cascade external cavity surface emitting laser (QC-VECSEL). The concept of the QC-VECSEL is to use THz QC-gain material to design a millimeter-scale reflective amplifying surface, or metasurface, for free space THz waves and incorporate it into a free-space THz resonant cavity to provide feedback to the amplification and form a laser. In this manner, the beam shape is determined by the external cavity, which supports fundamental Gaussian solutions. Further, the metasurface itself is composed of a subwavelength array (to prevent diffraction) of surface-coupled QC-elements whose properties, such as phase and polarization response, can be engineered on a unit cell basis allowing for a variety of unique experiments.
The power output power of the QC-VECSEL can be scaled by either increasing the size of the metasurface, or increasing the density (or fill factor) of QC-elements across the surface. In this work, large area metasurfaces with high fill-factor have been studied and demonstrated up to 1.35 W of peak output power for a QC-VECSEL operating at 3.4 THz at a heat sink temperature of 4 K. A peak wall-plug efficiency of ~2% is demonstrated, but observation of self lasing from the metasurface at high bias (when no external cavity is provided) in combination with a simultaneous roll-off in VECSEL output power suggests even higher efficiency can be achieved with improved suppression of self-lasing modes. The output beam is well fit to a Gaussian distribution with a 4 degree full-width half-maximum divergence angle.
In addition to power and beam quality, the QC-VECSEL opens the door to many interesting and unique studies via engineering of the metasurface properties and external cavity. Much of this thesis describes frequency tuning of QC-VECSELs based on broadband metasurfaces by varying the length of the external cavity. By making the external cavity extremely short (comparable to the operating wavelength), we are able to push all other external cavity modes outside of the gain bandwidth of the metasurface and demonstrate more than 20% fractional single-mode tuning around a center operating frequency of 3.5 THz. Because there are almost no diffraction losses at such a short cavity, the size of the metasurface could be reduced, allowing for continuous wave lasing with up to 20 milliwatts of output power at a heatsink temperature of 77 K, though the output power is highly variable as the reflectance of the output coupler has a strong frequency dependence. At the time of writing this, these are record performances in both frequency tuning and high-temperature continuous wave operation for lasers based on THz QC-gain material. The amount of tuning that be achieved with this approach is limited by the phase response of the metasurface, which squeezes the external cavity modes closer together in the spectral domain. Development of metasurfaces with lower electrical power consumption and higher conversion efficiency for the purpose of improving continuous wave performance. A sparse, patch-based metasurface with reduced power consumption is demonstrated, though the design was not optimal and only showed a 20% reduction in current draw compared to the previously demonstrated metasurfaces. Routes towards improving the performance are discussed.
The last subject discussed is the design of a mid-infrared (IR) QC-VECSEL. Due to the large metal losses at mid-IR frequencies compared to THz, the technique used to develop THz QC-VECSELs cannot be directly extended to the mid-IR. We propose a scheme based on a diffraction grating to provide surface coupling of the QC-gain material. Progress on experimental realization is discussed, but lasing has not yet been observed.