UC Santa Barbara

The invention of integrated circuits (ICs) in the 1960s and the subsequent microelectronics revolution fundamentally altered every aspect of modern technology, ranging from communication to computation and healthcare. Through the last few decades, continued miniaturization of such ICs has led to rapid improvements of device functionalities as well as exponentially reducing cost, thus making these technologies more accessible to everyone. This aggressive scaling has however, pushed device features towards sub-100 nanometer regimes, challenging conventional fabrication techniques. Semiconductor nanostructures can circumvent these challenges and enable further scaling for advanced logic and memory devices. Additionally, such nanostructures exhibit electronic and optical properties, that are of tremendous interest for alternative computing architectures such as quantum computing and photonic circuits, or to build power efficient and ultrafast platforms through low-energy and spin-based devices. Fabricating such nanostructures is not trivial but can be achieved through both “top-down” and “bottom-up” approaches. Top-down approaches, which are commonly followed in the semiconductor industry typically start with a bulk material and fabricate nanostructures through various etching steps. Although this technique is scalable and has been extremely successful in building modern ICs, the damage induced from etching can prove detrimental to the performance of low-dimensional systems. Bottom-up techniques generally refer to growing the nanostructures in an additive method in pre-defined positions on a wafer. Bottom-up approaches have the inherent advantage of exhibiting less defects, due to the elimination of etching steps and can be combined with in-situ patterning to build novel hybrid devices. However, these techniques are simultaneously more challenging to integrate and scale up reliably. For optimal performance of such nanostructures, it is critical to have precise control over their chemical composition, geometries, and material qualities. In this thesis, we will explore both bottom-up and top-down approaches, to achieve such defect-free scalable nanostructures in the context of low-power electronics and quantum computing. For bottom-up approaches, we investigate two templated epitaxial growth techniques: confined epitaxial lateral overgrowth (CELO) to grow III-V lateral heterostructures and selective area growth (SAG) to grow in-plane III-V nanowires coupled to superconductors. We first discuss the inherent advantages of CELO to fabricate low-power tunneling devices. Next, we explore the use of growth conditions to reduce defects, engineer facets and improve the material qualities and morphologies in CELO nanostructures. Using this, we demonstrate high-quality lateral III-V quantum wells for heterojunction tunnel transistors. For in-plane selective area growths, we investigate the effect of fundamental parameters such as growth temperature, cooldown processes and nanowire orientation and geometries on the nucleation of highly lattice mismatched heterostructures (indium arsenide on indium phosphide). These nanowires can lead to the fabrication of scalable systems with enhanced electrical and optical properties. We also develop in-situ shadowing techniques to create patterned heterostructures of dissimilar materials with pristine disorder-free interfaces. Subsequently, we use this to demonstrate superconductor-semiconductor hybrid nanowire networks for probing low-temperature effects in topologically non-trivial systems. In the last part of this dissertation, we will shift our focus to top-down techniques for defining aluminum/silicon/aluminum trilayer nanostructures with extreme aspect ratios. Such nanostructures can reduce footprint and enable scaling up of Josephson-junction based superconducting qubit systems. In conclusion, using both bottom-up and top-down techniques we combine advanced fabrication and epitaxial growth to achieve defect-free III-V scalable nanostructures with disorder-free interfaces. In the future, these nanostructures will enable the scalable fabrication of next-generation efficient computing platforms.