Refine
Year of publication
- 2011 (1) (remove)
-
Towards a THz Bloch laser
(2011)
- The realisation of tunable THz laser sources working at room temperature would give rise to further applications in this range of the electromagnetic spectrum. The THz Bloch laser could therefore become the basis for a technological breakthrough. Beside this practical relevance, the physics of the gain mechanism has been investigated theoretically for a long time and the experimental implementation of a self-starting laser still has not been achieved. At the beginning of this thesis the basic principles of Bloch oscillations and the related Bloch gain are described. The need of a superlattice structure to make Bloch oscillations possible in a semiconductor material is discussed. In this context, the effect of negative differential resistance and its influence on the field distribution due to Gunn domains is explained. The latter lead to an inhomogeneous field which may suppress the Bloch gain mechanism. The Krömer criterion is introduced and the concept of field-pinning layers to improve the field homogeneity is deduced. Finally, the design of the laser material is shown and different types of laser waveguides are compared. In chapter 3 detailed recipes for the processing of samples are given. Different types of contacts (ohmic and Schottky), the wafer bonding process required for double-metal lasers and the application of different photoresists for different purposes are described. An explanation of the formation of waveguides due to dry etching, wet etching and ion implantation follows. Dry etching is an established technique in the field of microstructure processing but the challenge of etching about 20 μm has led to problems. The high etching depth also makes wet etching difficult but this method could be improved due to a hard bake of the photoresist. The protection of critical areas on the surface of the samples with photoresist during ion implantation was increased by optimising the spin coating process. However, a full implantation of the active layer between the waveguides was not achieved which was the reason for the development of the hybrid technology. Here a prior wet etching of about 10 μm is performed and the rest of the material is implanted. The experimental setup is shown in chapter 4. An alternative method for the electrical contacting with the help of a copper bar is introduced. This improves the current distribution and the risk of an electrical breakdown during the measurements could therefore be lowered. Devices for THz beam guidance and spectroscopic measurements are shown and the method of biasing the samples with pulses below 100 ns and determining the effective voltage applied to the sample is depicted. These short pulses are required to prevent the samples heating up drastically due to high power. Chapter 5 contains the current-voltage characterisation of several structures including I-V-samples, Bloch laser samples and a quantum cascade laser. Different contacts (ohmic and Schottky) and different techniques for the formation of the ridges have been used in the processing of these samples (performed at the University of Frankfurt in all cases) and their influence on the I-V-dependence is discussed. The properties of the THz emission of the quantum cascade laser are in good agreement with published results from lasers processed with the same material. Another important result of this chapter is that the Bloch laser samples show unstable behaviour compared to the quantum cascade structure even with short pulses (of about 10 ns) where the risk of an electrical breakdown or the building of filaments is low. THz radiation emitted from one of the Bloch laser samples could not be observed. Two aspects that may have prevented the Bloch laser to emit are discussed in chapter 6. The saturation of the gain for higher amplitudes of the THz wave is investigated in single mode and multiple mode operation (the latter could occur due to the Bloch gain being expected to be broadband). In both cases it is shown that the saturation effect would limit the output power only to values clearly above the detection limit. In the subsequent section the distribution of the electric field is simulated with SILVACO software. Structures with transit layer lengths above the Krömer criterion are compared with structures which include field-pinning layers. It is shown that the latter are useful to avoid propagating Gunn domains as they build up in similar structures without field-pinning layers. Nevertheless, the electric field inside the superlattice regions is not stable. Beside spatial inhomogeneities also temporal variations of the field magnitude are observed. The lack of a suitable field distribution is expected to be the main reason for the samples not to work.
