Ultrafast spectroscopy of wide bandgap semiconductor nanostructures

Group III-nitrides have been considered a promising choice for the realization of optoelectronic devices since 1970. Since the first demonstration of the high-brightness blue light-emitting diodes (LEDs) by Shuji Nakamura and coworkers, the fabrication of highly efficient white LEDs has passed successful developments. A serious physical issue still remained, which prevents their use for high power and highly efficient LEDs: the drop of external quantum efficiency (EQE) of III-nitride LEDs when increasing the driving current, the so-called ''efficiency droop'' problem. In order to have a fast expansion to the lighting market, the cost-per-lumen of packaged LEDs must rapidly decrease. This indeed demands for having LED chips operating with high EQE under the high current operation. Besides the industrial interest of III-nitrides, owing to their large direct bandgap, they feature some interesting optical properties such as large exciton binding energy and large oscillator strength of excitons. However, when the carrier density raised in a semiconductor, a transition should occur from an insulating state consisting of a gas of excitons to a conductive electron-hole plasma, that is called the Mott transition. This crossover can drastically affect the optical and electrical characteristics of semiconductors and may, for instance, drive the transition from a polariton laser to a vertical cavity surface-emitting laser. More interestingly, even if biexcitons are frequently seen to dominate the emission of III-nitride heterostructures when the density is raised, no clear experimental report is available on the role of biexcitons in the Mott transition in a two-dimensional (2D) nanostructure. In the first part, the emission properties of high-quality GaN/AlGaN single quantum wells (QWs) at high-carrier densities are examined. They are of crucial importance to provide a deeper insight into the operating conditions of III-nitride based lasers and LEDs, as well as the transition from strong to weak coupling regime of exciton-polaritons in semiconductor microcavities. Employing the same technique then to investigate some InGaN/GaN QWs, the droop signature was investigated comprehensively in both polar and non-polar QWs. Having an accurate estimation of the carrier density in the QWs, the contribution of several non-radiative processes were quantitatively examined. These experiments can indeed provide a deeper insight on the physical phenomena responsible for the efficiency droop in III-nitride LEDs, and can stimulate several theoretical and experimental on this subject. The second part focuses on the transport mechanisms of excitons in ZnO and III-nitride based nano-structures. The transverse movement of donor-bound excitons in a purely bent ZnO microwire as a function of temperature have been investigated, owing to the high spatial, spectral and temporal resolutions of our original time resolved cathodoluminescence system. The movement mechanism was modeled by a hopping process of excitons. Our results pave the way to new experiments allowing to reveal the physics at play at the nanoscale in different materials. However, in case of highly disordered systems, one should take into account more complex considerations, as in the case of InGaN core-shell QW structures studied in the last chapter, the large energy fluctuations prevent excitons to move along the energy gradient at low temperatures.