Thesis presented November 08, 2022

Abstract:
Quantum computers rely on quantum two-level systems as memory units, the so-called qubits, and it is their quantum character what brings unique attributes that should enable to solve unaffordable problems with nowadays best classical computers. To this end, millions of qubits are required, and consequently large-scale quantum architectures capable of hosting and individually controlling them. Many quantum two-level systems have been proposed as qubits. Superconducting circuits have taken the lead and platforms with tens of qubits are already available at present. However, its scalability beyond a few hundreds is challenging. In this respect, quantum-dot-based spin qubits are among the latest to join the race, yet they seem to gather most of the requirements of a an ideal quantum processor. Current experiments deal with a few spin qubits at most, and one of the issues they may encounter when scaling up is variability. Spin qubits are sensible to their electrical environment, and defects in the surrounding materials may scatter their properties and yield to processors made of very heterogeneous qubits.
This thesis addresses the challenges that spin qubits may face in the near future both by understanding its impact and by proposing improved device candidates. We numerically quantify the variability for two of the most promising platforms for spin qubits: electrons and holes in Si MOS devices, and holes in Ge/SiGe heterostructures. We simulate their main sources of disorder, and discuss its repercussion on the realization of one- and two-qubit operations. We find that variability in Si MOS devices is a major challenge for scalability. We also evidence that variability, while smaller than for Si MOS, still shapes the properties of individual qubits in Ge/SiGe heterostructures, and propose a novel gate layout for Ge/SiGe-based processors that palliates its issues. Finally, we discuss how many-body interactions can reshape the physics of quantum dots, and how this must be accounted for in the design of spin qubit arrays.


Keywords:
Theory, Spin qubits, Quantum information, Modelling, Silicon, Germanium

On-line thesis.