Theoretical Study on Robust Control of Semiconductor Spin Qubits


Student thesis: Doctoral Thesis

View graph of relations


Related Research Unit(s)


Awarding Institution
Award date30 Aug 2018


In this thesis, we study several problems for the spin qubits based on semiconductor quantum-dot system in order to develop theoretical methodologies that facilitate the realization of high-fidelity quantum gates and reliable control of quantum systems. 

First, we begin with a brief introduction of the basic idea of quantum computation. We further introduce theoretical models describing the semiconductor spin qubits.

Second, we consider the case of a singlet-triplet qubit and study whether the barrier control is superior to the tilt control. Through randomized benchmarking simulations we have found that barrier control bears little to no advantage when the nuclear noise is significant if quantum gates are performed in traditional ways. We theoretically propose a new family of optimized pulse sequences and show that using these set of gates, the barrier control offers the substantial advantage over tilt control by extending the coherence time up to two orders of magnitude. These new optimized gates can reduce the gate times by up to 90% compared to traditional ones.

Third, we use the randomized benchmarking simulation method to theoretically study the responses of dynamically corrected gates for the exchange-only spin qubit under the 1/f noise environment which has noise spectra proportional to 1/ωα. We find that dynamically corrected gates can suppress hyperfine noises when α ≤ 1.5, but become inefficient otherwise.

Next, we study the microscopic properties of the triple-quantum-dot system and related quantum phenomena in order to understand the details in the qubit operation process. Particularly, we perform molecular-orbital-theoretic calculations to study the leakage in the system. We find that the leakage is 3-5 orders of magnitude smaller when the system is operated as a resonant-exchange qubit as compared to an exchange-only qubit. We also calculate the optimal detuning point at which the leakage is minimal. We find that although it is not identical to the double-sweet-spot at which the charge noise is substantially suppressed, they are quite close. We, therefore, conclude that the optimal strategy in manipulating a triple-quantum-dot system is operating at the double-sweet-spot point for charge noises as a resonant exchange qubit.

Finally, we summarize the results presented in this thesis and discuss possible future research directions in related fields.