Micro- and nanomagnet designs for the manipulation of spin qubits

This thesis discusses magnet designs that provide suitable stray fields for the manipulation of the spin direction of single electrons, enabling the implementation of a quantum computer. The building block, named qubit, is built by confining electrons in quantum dots with the quantum state encoded in the direction of each electron spin. The stray field allows for controlled rotations of the spin direction via an oscillatory displacement of the confinement potential minimum used to define the quantum dot. This technique, named electric dipole spin resonance (EDSR), has been implemented in several spin qubit processors with a Co micromagnet placed on top of the device. As a guideline, the requirements on the stray field of the magnet for successfully running quantum algorithms were set to a driving gradient > 1 mT/nm, a dephasing gradient < 0.1 mT/nm and single qubit addressability.
Extending this concept to a large number of qubits is challenging in terms of providing large enough driving gradients and individual addressability. Here we present three nanomagnet designs for this purpose, supported by micromagnetic simulations and the magnetic characterization of micro- and nanomagnets. We start by discussing the key parameters required in the modeling of magnets for EDSR manipulation. We demonstrate by scanning superconducting quantum interference device microscopy and simulations that the combination of polycrystallinity and large magnetocrystalline anisotropy energy inherent to Co micromagnets leads to random variations of the Larmor frequency of the qubits, negatively impacting qubit crosstalk and dephasing.
Building on these findings, we investigate the integration of dual-function Co magnetic structures within the qubit architecture. The magnets provide both the magnetic stray field for EDSR manipulation as well as the electric field modulation to confine the electron wavefunction. We fabricate a proof-of-concept architecture, demonstrating successful quantum confinement with Co nanogates. Moreover, we map their magnetization state in remanence by off-axis holography measurements. Micromagnetics simulations show that all the requirements of the stray field mentioned above are fullfilled with an applied external magnetic fields above 1 T.
Finally, we introduce two nanomagnet designs and discuss their advantages compared to micromagnets for the manipulation of spin qubits. The first design ("Horseshoe") is aimed at driving spin qubits arranged in linear chains and strongly confined in directions lateral to the chain. Spin-SEM measurements on arrays of such nanomagnets show reproducible magnetization patterns and good agreement with micromagnetic simulations. Based on these findings, we developed the design ("Block"), aimed at driving qubits arranged on a 2D lattice. Importantly, both designs are modular, ensuring efficient qubit manipulation independent of the number of qubits within the architecture.
These findings provide a novel concept for the development of a scalable electron spin qubit architecture manipulated with the stray field of magnets.