Growth and Doping Mechanisms of III-V Nanostructures by Selective Area Epitaxy
Selective area epitaxy (SAE), applied to semiconductor growth, allows tailored fabrication of intricate structures at the nanoscale with enhanced properties and functionalities. In the field of nanowires (NWs), it adds scalability by enabling the fabrication of well-ordered arrays, as well as networks. As a result, SAE-grown NWs are promising for a plethora of applications. At the beginning of this thesis, the method was not yet fully optimized, and numerous questions remained about the growth and fabrication. My research focuses on the homoepitaxial and heteroepitaxial growth of III-V horizontal nanostructures by molecular beam epitaxy (MBE) and metalorganic vapor-phase epitaxy (MOVPE).
We first investigate the effect of geometrical constraints on the SAE growth kinetics in MBE, taking GaAs nanomembranes (NMs) as a model system. Our study reveals the crucial role of adatom desorption and resorption from slit to slit as a function of width and pitch. We observe this effect both during the annealing and growth stages. Furthermore, better uniformity can be addressed using surfactant elements; Sb incorporation into the GaAs lattice promotes layer-by-layer growth. GaAs NMs are also used as templates for InAs NW growth. We correlated the NM faceting on different substrates with the NW morphology and the composition, which is critical for devising NW-based devices.
The functionality of the In(Ga)As NWs is strongly linked to the electrical properties, which requires the engineering of the carrier concentration and maximizing the carrier mobility. To this end, we investigated the technique of remote doping, during which dopants tend to diffuse within the structure. We thus turned to tackle the dopant segregation that tended to accumulate the dopants close to the interface of the InGaAs NWs on (111)B GaAs. By introducing a few ML-thick AlGaAs/GaAs layers, we suppressed the Si dopant diffusion by enhancing its solubility with Al. The study shows precise doping control in NWs, potentially enhancing device performance.
We then directed our efforts to integrate InP on Si(100) through molten alloy-driven liquid phase epitaxy to use these InP nanostructures as templates for earth-abundant Zn3P2 nanostructures. The idea behind this technique was to find conditions that lead to a single nucleus and thus, a single InP grain. We attempted to control the deposition of In droplets and the initial crystallization location by varying the temperatures and the amount of group-V flux. Our preliminary results on Zn3P2 growth on these templates are promising to replace InP-scarce substrates. Furthermore, In deposition within V-grooves resulted in horizontal metal NW formation, regardless of the length-width ratio. The optical properties of these In nanostructure arrays show UV plasmonic response.
My final work tackles the affordability aspect of substrate fabrication, a critical step for SAE. Employing UV and thermal nanoimprint lithography (NIL) to pattern nanosized openings can replace costly EBL patterning. We compare the two new approaches in terms of their scalability and uniformity.
Finally, I summarize the main findings of my research, giving an outlook for the future direction of this work. SAE is an essential approach for the epitaxial growth of nanostructures, and my work explains the influence of key parameters in SAE-grown nanostructures to achieve industrial scalability.
EPFL_TH10187.pdf
main document
restricted
N/A
82.34 MB
Adobe PDF
9e2978d8d3b3f9169ea658de26452f49