The quantum-cascade laser (QCL) emitting in the mid-infrared region of 4 to 8 m has been refined to the point that its internal quantum efficiency is approaching fundamental limits. QCLs designed for power typically contain 30-40 cascades, are less than two wavelengths in width, and laser ridge lengths are typically between 3 and 6 mm. Even with state-of-the-art efficiency and thermal management, room temperature operation of such lasers is fundamentally limited to several watts. This paper describes a path to power scaling that is not fundamentally limited. Power requires volume and thermal conductance. We propose that this combination is best achieved using fewer than 15 cascades combined with broad areas. We demonstrate the first room temperature continuous-wave emission of broad-area QCLs and discuss how this scaling concept can deliver MIR emission of 10's of watts at room temperature with beam quality required for high brilliance.
Active region designs of QCLs containing composite barriers based on AlAs allow short wavelength emission, improved injection efficiency, and high values of T0 and T1. On the other hand, AlAs introduces challenges, not only in strain compensation and growth, but also in effects on thermal management, thermal stability, and scattering. Leakage current, allowing electrons to bypass transitions between upper and lower laser levels occur due to scattering of electrons into higher-lying states via phonons and interface roughness scattering. This interface roughness scattering is exacerbated by large values of ΔEc and by the rms roughness itself, both of which are pronounced at the AlAs/InGaAs interface. The resulting leakage current noticeably reduces the slope efficiency, leading to more heating to achieve a given emission power. Efficient thermal management requires a buried heterostructure design; the re-growth of InP:Fe, however, needs to be carried out at temperatures consistent with maintaining the highly strained AlAs/InGaAs interfaces. This paper describes the physics of intersubband electron scattering due to strained interfaces and some partially optimized structures with Jth = 1.7 kA/cm2 at 300 K, slope efficiency η = 1.4 W/A, T0 = 175 K, and T1 = 550 K. Re-growth of InP:Fe using gas-source MBE at substrate temperatures below 550°C results in packaged lasers with 7 μm width having high thermal conductance.
We present progress on bandpass infrared interference filters with very narrow passbands to be used for sensitive trace gas and volatile compound imaging and detection and are suitable for mode selection and tuning in singlemode External Cavity Quantum Cascade Lasers. The process parameters for fabrication of such filters with central wavelengths in the 3-12 μm range are described. One representative fillter has a passband width of 6 nm or 0.14% with peak transmission of 62% and a central wavelength of 4.4μm. Theoretically, it can be tuned through about 4% by tilting with respect to the incident beam and offers orders of magnitude larger angular dispersion than diffraction gratings. We compare filters with single-cavity and coupled-cavity Fabry-Perot designs. The filters pass the tests for adhesion and abrasion as stated in MIL-C-48497.
The quantum-cascade lasers (QCL), first demonstrated in 1994, has since been developed into a mature laser emitting within nearly the entire spectrum from 2.6 to 250 μm, particular within the mid-infrared part of the spectrum from 3 to 12 μm for applications in gas sensing for security, environmental and medical uses, as well as for defense-related IR countermeasures. The QCL heterostructure is generally based on the InGaAs/InAlAs system lattice-matched to InP or on its strain-compensated extension to maximize the conduction band discontinuity between well and barrier material. A refinement is the use of mixed-height barriers to engineer the interface scattering of the different levels involved in the lasing process. This design strategy appears to be universally applicable, across the entire range of QCL emission wavelengths. By using low barriers where the upper laser state has its maximum probability and high barriers where the lower laser state has its maximal probability in strain-compensated designs for short wavelength emission, the lifetime of the upper laser state can be increased, while decreasing the lifetime of the lower laser state. First realizations of this design result in Jth = 1.7kA/cm2 at 300 K, slope efficiency η = 1.4 W/A, T0 = 175 K, and T1 = 550 K. Further increases in efficiency can be achieved through designs in which parasitic states near the upper laser level are separated from it, either energetically or oscillator strength. These states may be associated with other k values, or with higher-lying subbands.