MOVPE Growth Studies on Dilute Bismide Containing III/Vs & Development of an MOVPE In-Situ Gas Phase Analysis Setup
The strong rise of mobile and tethered data communication has a significant impact on global electricity consumption. Due to inefficient InyGa1 yAszP1 z telecommunication lasers (around 2 % efficiency with cooling efforts), 3 % of global electricity is consumed for optical data transfer. The low eff...
Keine Tags, Fügen Sie den ersten Tag hinzu!
|Zusammenfassung:||The strong rise of mobile and tethered data communication has a significant impact on global electricity consumption. Due to inefficient InyGa1 yAszP1 z telecommunication lasers (around 2 % efficiency with cooling efforts), 3 % of global electricity is consumed for optical data transfer. The low efficiency of those InyGa1 yAszP1 z telecommunication lasers is caused by loss processes, such as Auger recombination and IVBA, which lead to the heating of the devices. Other fields of III/V optoelectronics are also seeking more efficient candidates for device materials, like the search for 1 eV sub-cell alloys in multi-junction solar cells as well as a laser material on Si base. Different Bi containing III/V alloys are discussed as promising candidates. For example, GaAs1 xBix on GaAs with 10% Bi would be a highly efficient laser material for optical data transmission and provide a temperature insensitive band gap with 1.55 μm emission wavelength. The main reason for the high potential of III-Bi-V semiconductors is the fact that already small fractions of Bi, substituting group V host atoms, lead to a significant change of the band structure, which leads to suppression of Auger recombination and IVBA for sufficient compositions. However, the deposition of these highly metastable materials is challenging and still the subject of current research. In the present work, different dilute Bi containing III/V alloys were investigated. GaAs1 xBix, GaAs1 y xPyBix, GaAs1 y xNyBix, and GaP1 xBix were epitaxial grown using MOVPE and characterized using various structural and optical characterization techniques. The challenge here was to deposit structures with sufficient Bi fractions, while simultaneously realizing high quality layers and interfaces. The deposition of GaAs1 xBix with Bi fractions close to 10% has not been realized thus far. In the present work, the reasons for the Bi incorporation limit were under investigation. Therefore, alternative Bi MOs were used to unveil the influence of the thermal decomposition characteristics of the different precursors and the associated surface processes at the growth surface. It was shown that the incorporation limit is not dependent on the type of hydrocarbon residues (different precursors). Rather it was shown that the growth using different MOs led to nearly identical growth characteristics. Therefore it was concluded that at the temperature used neither the Bi incorporation limit is related to a specific hydrocarbon molecule at the growth surface, nor is an insufficient decomposition of one of the alternative MOs responsible. Quaternary layers GaAs1 y xPyBix and GaAs1 y xNyBix were deposited. They enable the lattice matched growth on GaAs, while simultaneously the band gap can be tuned independently over a wide range. GaAs1 y xPyBix was investigated as a potential candidate for a 1eV sub-cell material in multi-junction solar cells. It was possible to demonstrate the first PL activity in this quaternary alloy, which makes the material interesting for further optoelectronic applications. Furthermore, it was found that the smaller covalent radius of the P atoms led to an increased Bi incorporation limit. Hence, GaAs1 y xPyBix, GaAs1 y xNyBix, and GaAs1 xBix structures were deposited and compared. The two different quaternary materials in comparison to GaAs1 xBix showed that with increasing P (or N) incorporation the Bi incorporation limit was increased. Thereby, it was possible to prove the assumption that local strain is a crucial factor for the Bi incorporation limit. This is an important finding for future III-Bi-V studies, as it might open up the possibility of strain-engineering the Bi incorporation into III/Vs. Moreover, GaP1 xBix layers were deposited on GaP and GaP on Si. GaP1 xBix was a relatively new material, and the first deposition by MOVPE was successfully demonstrated. Furthermore, it was possible to incorporate high quality structures and Bi fractions up to 8.5 %. Originally considered to be a promising candidate for a laser material on Si, it was found that despite the large band gap reduction, the alloy is unlikely to lead to efficient light emitters. This is mainly related to a breakdown of the band edge Bloch character due to short-range alloy disorder and the indirect band gap. However, the findings are highly interesting for the Bi community from a theoretical point of view. Finally, it was possible to realize an improved GaAs1 xBix laser structure with an emission wavelength of 1015 nm at room temperature. The second project of this work was the development of a new in-situ setup of a MS connected to an MOVPE system. The new setup is meant to enable in-situ investigations of the deposition procedures discussed above. Especially of interest is the decomposition of MO precursors and analysis of MOVPE process desorption products. Therefore, with the support of Carl Zeiss SMT GmbH, a 3D ion trapped based MS prototype was prepared, and an in-situ setup to the MOVPE system was developed. The main challenge was to maintain the balance between transferring the analyte as unmodified as possible from the reactor chamber into the mass spectrometer and simultaneously not influencing the MOVPE process itself. Despite initial difficulties, the setup was successfully developed, and the potential was demonstrated by investigating the decomposition of TBAs. The main advantages of the new setup are the short measurement time for a mass spectrum over a large range and the ultra sensitive ionization conditions. The investigations on further MOs and growth processes are still under investigation at the moment of submission of this work and will be published separately. Altogether, it can be concluded that the investigations of this work led to various new insights into the growth of bismide containing III/V materials. Furthermore, the developed in-situ setup of the MS at an MOVPE system allows decomposition and growth investigations on a new level, which was demonstrated by TBAs decomposition experiments.|