The efficient integration of electronic and optoelectronic functions on a single chip within a CMOS-compatible material and technology platform is witnessing the next disruptive innovation in the microelectronics industry. Among the proposed approaches to “siliconize” photonics, the most ambitious is to develop a monolithic Si-based light source. The main challenge is the indirect bandgap of Si which hinders efficient light emission. A number of different solutions has been suggested to break down this material limitation, and the optimal strategy depends on the specific spectral range of application. In the THz range, quantum cascade laser (QCL) architectures based on the Group-IV material system have been predicted as viable potential sources up to room temperature1-3, by exploiting intersubband (ISB) transitions in the conduction band of n-type, Ge-rich Ge/SiGe multi-quantum well (MQW) heterostructures3. In this paper, we will show that the interface quality and threading dislocation density4 in this material system have finally reached the level required for possibly enabling, within a wide temperature range, a material gain larger than the cavity losses5. From the structural standpoint, we report on the growth by ultra-high-vacuum chemical vapor deposition (UHV-CVD) of strain-compensated QCL stacks with a thickness up to 10 µm, demonstrating by X-ray diffraction and scanning transmission electron microscopy their high crystalline quality and remarkable growth reproducibility [Fig. 1(a)]. We will also describe optimized designs obtained in waveguide modelling with which waveguide losses comparable to III-V architectures are achieved6. In the model system of n-type Ge-rich asymmetric coupled quantum wells (ACQWs)7,8, being the building block of a QCL structure, we will prove a high degree of control on the engineering at the nanoscale the intersubband electronic spectrum and the wavefunctions relevant for tunneling processes, which allowed us the observation of photoluminescence at 4 THz after optical excitation using a free-electron laser (FEL)9. In addition, we will present a systematic calibration study of the material parameters controlling the interface roughness (IFR) and electron-phonon scatterings in n-type Ge/SiGe MQWs by combining experimental pump-probe data with a numerical model of ISB carrier dynamics including inelastic and elastic scattering channels. The time evolution of subband populations after FEL optical pumping is investigated as a function of the MQW design geometry, comparing symmetric and stepwise configurations, and analyzed by varying the subband energy spacing changing the well width. As a matter of fact, these material parameters critically affect the outcomes of the non-equilibrium Green’s function calculations widely used for predicting quantum transport and, ultimately optical gain of QCL devices10,11 and so to guide the optimization of proposed device designs. Hitherto, however, in most cases, the adopted values for these material parameters were not calibrated against experiments. For a proper calibration of the deformation potentials, we studied the non-radiative lifetimes as a function of the subband spacing on a set of symmetric MQWs of different Ge well widths. The corresponding α2D absorption spectra are reported in the left panels Fig. 1(b-d). Differential transmission spectra from single-color pump-probe experiments and model simulations indicate [Fig. 1(e-g)] that the electron-phonon coupling is much less effective than that expected from the bulk deformation potentials reported by Jacoboni et al.12. In particular, we found that the relaxation mediated by the emission of intervalley OPs is suppressed in low-dimensional multilayer structures while the intravalley OP scattering is described by a reduced effective deformation potential, thus explaining the observation of relaxation times larger than 10 ps above both the OP energy thresholds in single-color pump probe spectroscopy experiments. These results are crucial for developing adequate simulation frameworks and specific design approaches for achieving optical gain in silicon-compatible quantum cascade emitters.[1] K. Driscoll et al., Appl. Phys. Lett. 89, 191110 (2006).[2] K. Driscoll et al., J. Appl. Phys. 102, 093103 (2007).[3] D. J. Paul, Laser & Photonics Reviews 4, 610 (2010).[4] T. Grange et al., Physical Review Applied 13, 044062 (2020).[5] T. Grange et al., Appl. Phys. Lett. 114, 111102 (2019).[6] K. Gallacher et al., Opt. Express 28, 4786 (2020).[7] C. Ciano et al., Physical Review Applied 11, 014003 (2019).[8] L. Persichetti et al., Crystals 10, 179 (2020).[9] C. Ciano et al., Opt. Express 28, 7245 (2020).[10] A. Valavanis et al., Phys. Rev. B 83, 195321 (2011).[11] C. Jirauschek et al., Applied Physics Reviews 1, 011307 (2014).[12] C. Jacoboni et al., Phys. Rev. B 24, 1014 (1981). Figure 1
Read full abstract