Capture, relaxation and recombination in quantum dots

Quantum dots (QDs) have attracted a lot of interest both from application and fundamental physics point of view. A semiconductor quantum dot features discrete atomiclike energy levels, despite the fact that it contains many atoms within its surroundings. The discrete energy levels give rise to very narrow optical emission lines at low temperature. When exciting an ensemble of quantum dots, the discrete emission lines are however broadened by the size distribution of the quantum dots. In particular, the advancement of different growth techniques like Molecular Beam Epitaxy and Metal Organic Vapour Phase Epitaxy has lead to the development of different types of ultrapure QD nanostructures with well-defined size, shape and composition. Nowadays QDs can even be grown in laterally ordered patterns. In the last decade, great progress has been obtained in the detailed physical understanding of these semiconductor nano-structures. A complete understanding of the carrier dynamics in QDs, which is essential for exploiting the utilization potential predicted for these QDs, is nevertheless lacking. In this thesis, the carrier dynamics of different types of InAs/GaAs QDs is investigated. When the carriers are photo-generated in the barrier layer much above the QD bandgap, they diffuse towards the QDs, in which they are captured and subsequently undergo relaxation towards the QD ground state. After the relaxation, each electron-hole pair within the QD ground-state with total spin J=±1, recombines radiatively and emits a single photon with an energy corresponding to the separation of the confined energy level within the QD. In this Thesis, QD-materials in which many defects are intentionally included by lowtemperature growth, is also investigated. In this case, the photo-generated carriers in the bulk GaAs-barrier and in the QDs can also be nonradiatively trapped into nearby defect states. An overview of the carrier capture, relaxation and trapping routes in both high quality QD structures and in low temperature grown QD structures is presented in Chapter 2. In this Thesis, the carrier dynamics within III-V semiconductor QDs are studied using the Time Resolved Differential Reflectivity (TRDR) and Photoluminescence (PL) measurement techniques. The TRDR experiment employs a two-color pump-probe system, in which the pump laser has a photon energy tuned above the GaAs barrier bandgap. The optically generated carriers are captured into the QDs inducing a refractive index change of the QDs. The resulting change of the reflectivity of the probe beam, which is tuned in resonance with the QD ground state, is subsequently detected as a function of the pump-probe delay to obtain the carrier dynamics. The reflectivity is interpreted under the assumption that the QD layer can be considered as a thin continuous layer with an effective dielectric constant. In this case, the observed reflectivity signal is basically the interference between the surface reflection and the small reflection from the QD-layer. The advantage of TRDR technique lies in the fact that it is non-destructive since it does not require etching of the substrate. Secondly, the TRDR technique allows the study of the QD-samples not only at low temperature but also up to room temperature. Finally, the TRDR technique allows the study of samples in which the non-radiative recombination is (intentionally) very high and in which the PL efficiency is thus too small to be measured. In Chapter 3, an array of InAs/GaAs QDs grown on a quantum wire (QWR) superlattice (SL) template is investigated to understand the capture and relaxation mechanism in these QDs. This particular QD structure is chosen since it is expected to reduce the carrier diffusion from the bottom barrier layer into the QDs, allowing a more controlled study of the actual carrier capture mechanism. The quantum wire superlattice is spectrally well separated from the QDs, thereby also allowing the study of re-distribution of carriers from the QWR template into the QDs. First, the PL spectrum from this sample is studied and observed a clear carrier re-distribution from the QWR template to the QDs at temperatures between 40 K and 100 K. Subsequently, TRDR experiments were performed, showing a distinct increase of the TRDR rise time between 40 K and 100 K. TRDR experiments as a function of the pump excitation density at 5 K showed a TRDR risetime which decreases from 28 ps at a low excitation density to 5 ps for high excitation density, also showing a plateau for excitation densities above 1 kW cm-2. At room temperature, the TRDR risetime is apparently independent of the excitation density. These observations make it clear that a simple Auger or phonon relaxation picture is not capable of explaining the carrier capture and relaxation in this sample. The data presented in this Thesis can be explained by taking into account the carrier diffusion through the barrier layers towards the capture volumes of the QDs. The diffusion through the barrier is modified by the capture and re-emission of carriers from the superlattice template. At low excitation density, the number of carriers excited directly within the QD capture volume is less than unity, implying that the carrier have to first diffuse through the barrier towards the QDs, thereby resulting in a longer risetime. As the excitation density is increased, more carriers are excited directly within the capture volume, thereby decreasing the risetime. At low temperature, the carriers generated in the bottom barrier layer are all captured by SL template, thus reducing the effective diffusion length. When the temperature increases, carrier re-emission from the SL template increases the effective carrier diffusion length, thus explaining the observed increase of the TRDR risetime. In Chapter 4, a study of an InAs/GaAs Stransky-Krastanow grown quantum rod (QR) ensemble is presented, which have been grown by the leveling and rebuilding technique. In these structures the TRDR decay dynamics was investigated, which was found to be modified by the number of resonant QDs probed within the total QR size distribution. It is found that the TRDR decay time develops a clear resonance at high excitation densities, while at low excitation density the well-known excitonic behavior is regained. At very high excitation density, and also at high temperature, the resonant enhancement of the TRDR decay time gradually decreases. These results are interpreted in terms of an electromagnetic coupling of the excited QRs. Due to this coupling, the QRs collectively decay, giving rise to a polaritonlike picture. The experimental results can be interpreted in a picture in which the TRDR lifetime of the QRs is determined by the separation between identical and excited QRs, which are in resonance with the probe. At low excitation density, the density of excited QRs is too small for such a collective decay behavior, leading to the excitonic life time. At high temperature and also at very high excitation density, the exciton dephasing rate becomes much larger than the radiative decay rate, which washes out the electromagnetic coupling. Chapter 5 presents the investigation of low temperature (250 °C) grown InAs/GaAs QDs which have been subjected to different annealing conditions. The purpose is to investigate the low temperature (LT) grown QDs as a potential device material to be used in ultra-fast optical switching. It was found that the PL-efficiency of the QDs is reasonable at low temperature, but rapidly quenches with increasing temperature and disappears at 40 K. It is also observed that the PL efficiency substantially increases when the LT-QDs are excited below the GaAs bandgap, directly into the QDs. These results indicate that LT growth indeed provides many anti-site defects within the GaAs barrier which act as fast trapping centers for the photo-excited carriers, but the QDs itself are free from these PL quenching centers. Finally, in Chapter 6, the decay dynamics of the LT-QDs is studied using the TRDR technique, since the TRDR technique is expected to be still sensitive at room temperature and for samples in which the PL is quenched by excess nonradiative recombination. From the rapid quenching of the PL-efficiency between 5 and 40 K observed in Chapter 5, a strong decrease of the TRDR decay time was expected for these LT-QDs. The TRDR decay time is however observed to be slow. Quite surprisingly, the TRDR decay time is measured to beindependent of the temperature. These data are interpreted by noting that the photo-excited electrons in the GaAs barrier are very efficiently trapped by the charged anti-site defects, while the holes are less efficiently trapped by neutral anti-site defects. As a result, the holes will be captured within the LT-QDs, while the electrons will mainly remain within the traps in the GaAs barrier. The TRDR signal is proportional to the sum of the electron and hole occupation within the LT-QDs and will thus mainly probe the hole dynamics which is not expected to be ultra-fast. The PL technique probes the product of the electron and hole densities, and thus effectively probes the very small electron occupation within the QDs. The TRDR decay time is interpreted as the trapping recombination time between holes confined within the LT-QDs and the neutral anti-site defect within the GaAs barrier, in close proximity to the LT-QD.