(111) Si spin qubits constructed on L point of band structure

Abstract

(001) Si spin qubits are being intensively studied because they have structures similar to that of CMOS devices currently being produced, and thus have the advantage of utilizing state-of-the-art miniaturization, integration, and variation-reduction-technologies. However, there are still issues, such as further improvement of relaxation and decoherence time, stabilization of valley-splitting control, and reduction of the variation caused by the roughness of the interface. In this study, new measures are proposed to address these three issues. Instead of confining an electron to the minimum energy point $X_0$ of the conduction band along the band structure $\Gamma-\Delta-X$ in (001) Si crystals, we propose confining an electron to the L point along $\Gamma-\Lambda-L$ in (111) Si crystals. At the $X_0$ point, the symmetry causes spin-orbit interaction to act on the electron, and the sixfold degeneracy is lifted into a fourfold and a twofold, and the valley-splitting of the twofold conflicts with the two-level system. In the symmetry of the $L$ point, substantial spin polarization disappears, facilitating the reduction of the spin-orbit interaction, and the fourfold degeneracy is lifted to threefold and single, and the single becomes ground state. The need to increase the magnitude of the valley-splitting is exempt, allowing the electron to be controlled away from the interface, which is expected to reduce the variation caused by the roughness of the interface. Data on the confinement of an electron to the $L$ point and the control of fourfold degeneracy are needed, and it is hoped that prototype silicon spin qubits constructed on (111) Si crystals will be developed and that the proposed device structure will help implement quantum computers based on Si devices.