Theory on Robust Manipulation of Silicon-based Spin Qubits

Description

A quantum computer is expected to outperform the current information technology inmany important problems. One of the most promising candidates for its realization is thesemiconductor spin qubit. Compared to other materials, silicon-based quantum deviceshold a unique advantage that one of the isotope of silicon, the silicon-28, carries nonuclear spin. Qubits built on isotopically enriched silicon essentially live in a "spinvacuum", and decoherence due to nuclear spin bath is greatly suppressed. With thisadvantage, it has been experimentally demonstrated that one may control a qubit basedon isotopically enriched silicon with extremely long coherence time and very high controlfidelity.Nevertheless, it remains a challenge to further suppress the gate error. The charge noise,caused by unintentionally deposited impurities near the quantum dots, is the criticalcontributing factor to the decoherence of a qubit based on isotopically enriched silicon.In order to combat the noise, Dynamically Corrected Gates have been developed. Amongvarious proposals, the SUPCODE is particularly useful for the singlet-triplet qubit whichuses the two-electron singlet and triplet states as the computational bases. SUPCODEhas been optimized for isotopically enriched silicon quantum device: a kind ofSUPCODE which treats charge noise alone are not only found to be about 40% shorterthan the full SUPCODE but also offers better performances. This and many otherpreliminary results as detailed in the proposal suggest that it is possible to optimize thequantum gates in silicon-based semiconductor spin qubits.In this project, we will develop efficient, optimized protocols to control spin qubits basedon silicon. We will start with an investigation on the change noise, in which we will focuson quantitatively how the charge noise changes under different controlling schemes.Based on these results, we will proceed to pulse design and optimization. We will usenonzero exchange interaction to shorten the pulse sequences that have been developed,employ the technique of smooth pulse shaping in order to develop optimized,experiment-friendly pulses that improve the performances of both single and two-qubitgates. We will also extend the study to multi-qubit gates, to devices with multipleelectrons, and to the control problem in a rotating frame. Overall, the results of thisresearch are expected to provide a viable means to control silicon spin qubits, whichshould constitute an important step forward in accomplishing fault-tolerant, scalablequantum computation using silicon quantum devices.