Epitaxial Growth of Highly Mismatched Alloys on GaAs Substrates

Global warming mandates change of our energy budget and sources towards renewable energies8,9. The internet has a total consumption of 10 % of the world's electricity demand increasing every year10. Poor efficiency of the InP laser device used for optical fiber communication is one particular issue12–15. More efficient laser devices on the GaAs platform would provide a long-desired alternative. In the future, the present thesis may give a contribution to reduce energy consumption in fiber communication systems. Based on dilute bismides two approaches have been investigated for laser device application on GaAs substrates. The growth of (Ga,In)(As,Bi) on GaAs revealed a limitation on the accessible compressive lattice mismatch. Instead of increasing the compressive lattice mismatch of the layer, the Bi fraction has been reduced upon In incorporation at a constant level of lattice mismatch. The reduced Bi incorporation has led to surplus Bi segregating to the growth surface, eventually forming metallic droplets. Hence, the desired redshift of the emission wavelength compared to ternary Ga(As,Bi) has not been possible. Future experiments should focus on a deeper understanding of the Ga(As,Bi) growth, especially at lower temperatures to increase the Bi fraction before alloying with another material. Development of alternative precursors for low-temperature metallorganic vapor phase epitaxy growth can help to achieve the desired composition. The growth of the Type-II heterostructures gives a promising path for further investigations towards laser emission at 1.55 µm on GaAs substrates. A crucial challenge is the interface formation during growth in these heterostructures, which requires significant adaptions to the growth conditions at the interfaces. To optimize the optical quality of the Ga(As,Bi) layer a Bi wetting of 60 s has been established to account for segregation effects. In future experiments, the Bi wetting needs to be optimized on Ga(N,As) layers to ensure a high optical quality of WQW heterostructures even at high N fractions. At the second interface of a WQW the surplus Bi must be desorbed before the growth of the second Ga(N,As) layer. Otherwise, the N incorporation is suppressed. Therefore, a temperature increase to 625 °C has been established with subsequent annealing for 120 s to allow all residual Bi to desorb. For future experiments, the molar fractions and thicknesses of the respective layers should be optimized for particular emission wavelengths, namely 1.55 µm and beyond. Challenge will be the optimization of the Bi wetting, which could be performed in-situ using RAS. Room-temperature laser operation of (Ga,In)As/Ga(As,Bi)/(Ga,In)As WQW laser device is a thriving breakthrough for Ga(As,Bi) based Type-II heterostructures, proving the concept and justifying further research. An emission wavelength of 1037 nm has been achieved with an optical efficiency of 48 mW/A. Starting with this prototype structure, the emission wavelength can be shifted further by carefully adjusting the molar fractions and thicknesses. Even though no laser device with an emission wavelength of the technologically essential wavelengths of 1.3 µm and 1.55 µm has been demonstrated, the present thesis gives promising paths for Ga(As,Bi) based Type-II heterostructures in achieving the long-desired goal for long-wavelength lasers on GaAs substrates.