- We explain an unexpected anomoly of the g-factor (how the electron spin qubit energy changes as a function of magnetic field) in silicon quantum dots.
- This new mechanism creates a new channel for decoherence (via electrical noise and spin-orbit coupling), which – we predict can be negated by changing the direction of the magnetic field.
- This new tunability of the g-factor can also be used to make 2-dot encoded spin qubits with full Bloch sphere control (without the need for a magnetic gradient).
We theoretically model the spin-orbit interaction in silicon quantum dot devices, relevant for quantum computation and spintronics. Our model is based on a modified effective mass approach with spin-valley boundary conditions, derived from the interface symmetry under presence of perpendicular to the interface electric field. The g-factor renormalization in the two lowest valley states is explained by the interface-induced spin-orbit 2D (3D) interaction, favoring intervalley spin-flip tunneling over intravalley processes. We show that the quantum dot level structure makes only negligible higher order effects to the g-factor. We calculate the g-factor as a function of the magnetic field direction, which is sensitive to the interface symmetry. We identify spin-qubit dephasing sweet spots at certain directions of the magnetic field, where the g-factor renormalization is zeroed: these include perpendicular to the interface magnetic field, and also in-plain directions, the latter being defined by the interface-induced spin-orbit constants. The g-factor dependence on electric field opens the possibility for fast all-electric manipulation of an encoded, few electron spin-qubit, without the need of a nanomagnet or a nuclear spin-background. Our approach of an almost fully analytic theory allows for a deeper physical understanding of the importance of spin-orbit coupling to silicon spin qubits.