Carrier dynamics in GaAs-based photoconductive terahertz (THz) sources is investigated using Monte Carlo techniques to optimize the emitted THz transients. A self-consistent Monte Carlo-Poisson solver is developed for the spatio-temporal carrier transport properties. The screening contributions to the THz radiation associated with the Coulomb and radiation fields are obtained self-consistently by incorporating the three-dimensional Maxwell equations into the solver. In addition, the enhancement of THz emission by a large trap-enhance field (TEF) near the anode in semi-insulating (SI) photoconductors is investigated. The transport properties of the photoexcited carriers in photoconductive THz sources depend markedly on the initial spatial distribution of those carriers. Thus, considerable control of the emitted THz spectrum can be attained by judiciously choosing the optical excitation spot shape on the photoconductor, since the carrier dynamics that provide the source of the THz radiation are strongly affected by the ensuing screenings. The screening contributions due to the Coulomb and radiation parts of the electromagnetic field acting back on the carrier dynamics are distinguished. The dominant component of the screening field crosses over at an excitation aperture size with full width at half maximum (FWHM) of ~100 um for a range of reasonable excitation levels. In addition, the key mechanisms responsible for the TEF near the anode of SI photoconductors are elucidated in detail. For a given optical excitation power, an enhancement of THz radiation power can be obtained using a maximally broadened excitation aperture in the TEF area elongated along the anode due to the reduction in the Coulomb and radiation screening of the TEF.

The photoconductive method is one of the oldest methods for the generation of THz room temperature operated THz electromagnetic waves. The THz photoconductive source has operated at a lower power level in the order of hundreds of nW. In addition, the energy conversion of optical to THz efficiency has remained extremely low. One of the most efficient THz photoconductive sources is a trap-enhanced field (TEF) effect source. The field is measured to contain more than 90% of the total DC bias within the first 5 µm of an 80 µm gap between the electrodes reaching kV/cm with only a modest bias. The overall THz power, however, has remained low, due to its rapid saturation. To date, there has been a limited understanding of the TEF effect. In this thesis, a more detailed experimental investigation of TEF effect current transport and field distribution based on annealing is presented to explain some of the underlining physics of TEF effect. A spatially extended line excitation is introduced to effectively reduce the screening effect while still exploiting the TEF region to maintain high efficiency and reach the µW regime. The record efficiency reached by this method is demonstrated. An experimental demonstration with a numerical analysis of the line excitation is presented. The spectral analysis of both a point and a line excitation demonstrate that the line excitation spectrum is not only comparable to that of the point excitation, but also extends the range of useful lower frequency content. To further improve the THz efficiency, the line excitation THz array is investigated.

Ausgehend von einem thermodynamisch konsistenten Modell von Dreyer und Duderstadt für Tropfenbildung in Galliumarsenid-Kristallen, das Oberflächenspannung und Spannungen im Kristall berücksichtigt, stellen wir zwei mathematische Modelle zur Evolution der Größe flüssiger Tropfen in Kristallen auf. Das erste Modell behandelt das Regime diffusionskontrollierter Interface-Bewegung, während das zweite Modell das Regime Interface-kontrollierter Bewegung des Interface behandelt. Unsere Modellierung berücksichtigt die Erhaltung von Masse und Substanz. Diese Modelle verallgemeinern das wohlbekannte Mullins-Sekerka-Modell für die Ostwald-Reifung. Wir konzentrieren uns auf arsenreiche kugelförmige Tropfen in einem Galliumarsenid-Kristall. Tropfen können mit der Zeit schrumpfen bzw. wachsen, die Tropfenmittelpunkte sind jedoch fixiert. Die Flüssigkeit wird als homogen im Raum angenommen. Aufgrund verschiedener Skalen für typische Distanzen zwischen Tropfen und typischen Radien der flüssigen Tropfen können wir formal so genannte Mean-Field-Modelle herleiten. Für ein Modell im diffusionskontrollierten Regime beweisen wir den Grenzübergang mit Homogenisierungstechniken unter plausiblen Annahmen. Diese Mean-Field-Modelle verallgemeinern das Lifshitz-Slyozov-Wagner-Modell, welches rigoros aus dem Mullins-Sekerka-Modell hergeleitet werden kann, siehe Niethammer et al., und gut verstanden ist. Mean-Field-Modelle beschreiben die wichtigsten Eigenschaften unseres Systems und sind gut für Numerik und für weitere Analysis geeignet. Wir bestimmen mögliche Gleichgewichte und diskutieren deren Stabilität. Numerische Resultate legen nahe, wann welches der beiden Regimes gut zur experimentellen Situation passen könnte. / Based on a thermodynamically consistent model for precipitation in gallium arsenide crystals including surface tension and bulk stresses by Dreyer and Duderstadt, we propose two different mathematical models to describe the size evolution of liquid droplets in a crystalline solid. The first model treats the diffusion-controlled regime of interface motion, while the second model is concerned with the interface-controlled regime of interface motion. Our models take care of conservation of mass and substance. These models generalise the well-known Mullins-Sekerka model for Ostwald ripening. We concentrate on arsenic-rich liquid spherical droplets in a gallium arsenide crystal. Droplets can shrink or grow with time but the centres of droplets remain fixed. The liquid is assumed to be homogeneous in space. Due to different scales for typical distances between droplets and typical radii of liquid droplets we can derive formally so-called mean field models. For a model in the diffusion-controlled regime we prove this limit by homogenisation techniques under plausible assumptions. These mean field models generalise the Lifshitz-Slyozov-Wagner model, which can be derived from the Mullins-Sekerka model rigorously, see Niethammer et al., and is well-understood. Mean field models capture the main properties of our system and are well adapted for numerics and further analysis. We determine possible equilibria and discuss their stability. Numerical evidence suggests in which case which one of the two regimes might be appropriate to the experimental situation.