Dopant Qubits in Direct Band Gap Materials

Abstract

Quantum information is currently one of the most important research fields in physics. The dopant qubits in direct band gap materials are promising candidates to build large- scale photon-based quantum network for applications in quantum computation and quantum communication. In this thesis, we study the optical and spin properties of different dopant systems in order to find the best one as a qubit candidate. The first study of dopant qubits in direct band gap materials was on donors in GaAs, from Yamamoto’s group about 10 years ago. The donor qubits in GaAs only have ns-scale dephasing time which is limited by inhomogeneous hyperfine fields in the crystal. In order to improve the coherence time, we study the acceptors in GaAs which use the hole spins as the qubit states instead of the electron spins for donors. As the hole has a Bloch wave function with p-symmetry, the overlap with nuclear spin is zero thus the hyperfine interaction is weaker. However, from our measurements, we find the dephasing time is still in ns-scale. This short dephasing time is because of the long-range dipole-dipole hyperfine interaction together with a non-zero heavy-hole-light-hole mixing. As the acceptor system does not show specific advantages, we move back to the donor system and study the system in different materials. We have performed a systematic study of the longitudinal spin relaxation time T1 for donor qubits in three different materials (GaAs, InP, and CdTe). For these three materials, the spin relaxation times are all in ms-scale and our theory study shows the T1 is limited by phonon interaction mediated by spin-orbit coupling. Though none of these three materials show great coherence properties, this T1 study leads to our research on donors in ZnO which is a material with much weaker spin-orbit coupling thus longer T1. We have demonstrated optical spin initialization and coherent control for donors in ZnO. With these spin-control techniques, we have measured a T1 up to hundreds of ms, an inhomogeneous dephasing time T2* of 17 ns and a spin-echo dephasing time T2 of 50 μs. The donors in ZnO have T1 several orders of magnitudes higher than the other dopant systems we have studied. Even though the T2* is only in ns-scale, ZnO can potentially get nuclear-spin-free isotopic purification, which can greatly enhance the dephasing time. To utilize donors in ZnO for the quantum network applications, it is necessary to get single donor and high-fidelity spin control. We show our current progress on single donor isolation with focused ion beam and high-fidelity spin control with microwave. Though these two tasks are not accomplished yet, the progress is promising and we expect they can be achieved in the next few years.