As microelectronic components get smaller and smaller, the influence of the metallic interconnects on the processing speed and on the energy consumption increases dramatically. To circumvent the interconnect bottleneck, the metallic wirings on a Si-chip could be replaced by optical interconnects. The desired compatibility to already established CMOS-technology led to the rise of Silicon Photonics. The biggest challenge and \enquote However, their optical properties never matched the expectations due to the spatial separation of electrons and holes and the indirect nature of the bandgap of both constituents. One way to enhance the spontaneous emission rate is to embed SiGe QDs in photonic crystal cavities, taking advantage of the Purcell effect. In the first part of this work, micro-photoluminescence (-PL) measurements on single SiGe QDs embedded in L3 photonic crystal cavities are discussed. The relative position of the QD to the cavity is systematically changed. As predicted by the Purcell effect, the emission rate depends heavily on the spatial overlap of the emitter and the cavity mode. With this approach, the local density of states was mapped for individual L3 cavity modes. As suggested by M. Schatzl in her PhD Thesis, the single exponential decay at low excitation powers of single SiGe QDs embedded in an H1 cavity could be a sign of single photon emission. Hanbury-Brown and Twiss experiments were conducted to validate this statement. However, no anti bunching was observed.
The second part of this work deals with defect engineering of Ge QDs; a novel approach for the realization of Si-based light emitters. Single defects are introduced into the QDs by Ge ion implantation during their growth and subsequent annealing. These defects lead to deep, localized electron states below the conduction band edge of Ge. Optically direct transitions are observed up to room temperature (RT) by recombination of electrons bound to those states with holes confined inside the QDs. These defect-enhanced Ge quantum dots (DEQDs) can be used as active gain material for CMOS-compatible laser sources. Here, various approaches to improve the RT PL yield of DEQDs are presented. Both in-situ and post-growth sample treatments are discussed. In-situ treatment includes the influence of doping and thus, the increase of charge carriers in the layer system. For an ideal Sb doping concentration, we observe a PL enhancement of more than a factor of two, due to the supply of additional electrons. The influence of the ion implantation species into the QDs (e.g. using Si instead of Ge as implanted species) is discussed. An increase of PL yield is observed if the implantation conditions for Si ions are chosen properly. Furthermore, we will discuss the influence of post-growth treatment on the PL yield of DEQDs, i.e. annealing, hydrogen implantation, and passivation. Combining the aforementioned approaches leads to an optimization of DEQDs as gain material for electrically pumped CMOS-compatible lasers operating at RT.