PhD Defense: Martin Bjergfelt

 In-situ patterned superconductor/semiconductor nanowires for quantum devices

The quality of interfaces and surfaces is crucial for the performance of nanoscale devices. A pertinent example is the close tie between current progress in gate-tunable and topological superconductivity using superconductor/semiconductor electronic devices and the hard proximity-induced superconducting gap obtained from epitaxial aluminium/indium arsenide heterostructures. Fabrication of devices requires selectively etching superconductor segments from the semiconductor; this is currently only possible with Al/InAs, curbing the functional use of other superconductor/semiconductor material combinations.

In this thesis defense, I present a new crystal growth platform based on three-dimensional structuring of growth substrates which is independent of the choice of semiconductor and superconductor, and enables the synthesis of semiconductor nanowires with in-situ patterned superconductor shells. This wafer-scale technique enables realisation of all the most frequently used architectures in superconducting hybrid devices. Characterisation using electron transport at cryogenic temperatures revealed an increased device yield, electrostatic stability, along with evidence of ballistic superconductivity compared to etch-processed devices. In addition to aluminium, indium arsenide nanowires with patterned indium, vanadium, tantalum and niobium shells are presented. 

Structural characterisation of the hybrid nanowires by transmission electron microscopy was used to correlate the crystallinity of the superconductor shells to the superconducting transport properties. A range of crystallinity was observed, from epitaxial crystals in aluminium and indium to amorphous or nanocrystalline films of vanadium, tantalum and niobium. The occurrence of these morphologies can be understood through consideration of the relationship between nucleation probability, surface energies and interfacial strain. Using a wide range of materials increases understanding of superconducting properties of materials at the nanoscale, as well as the conditions for obtaining hard-gap proximity-induced superconductivity in hybrid devices. Overall, this work shows that long-range epitaxy between a superconductor and a semiconductor is not a prerequisite for proximity-inducing a hard superconducting gap in all superconductors. Impurity-free, uniform interfaces may be sufficient if the thin film itself supports well-defined superconductivity. The presented platform opens for future work on advanced device architectures, new combinations of hybrid nanomaterials and in-situ protection of surfaces.