Download files
Abstract
A quantum computer utilizes the laws of quantum mechanics and performs logic operations on quantum two-level systems, known as quantum bits, or qubits. In this PhD thesis, the qubit is defined by the spin states of a single electron confined in a silicon metal-oxide-semiconductor (MOS) quantum dot, a platform that is highly compatible with the fabrication technology currently employed in today's semiconductor industry.
We have realized a MOS quantum dot qubit using an isotopically-purified silicon substrate, with an on-chip single-electron transistor (SET) for spin readout and a broadband microwave line for delivering electron spin resonance (ESR) control pulses. The qubit has a coherence time of 28 ms and a control-fidelity of 99.6%, which is above the threshold required for fault-tolerant quantum computing. In addition, a gate voltage can be used to Stark shift the qubit's resonance frequency over 3000 times the minimum ESR linewidth, hence opening up the prospect of independent qubit addressability in a large-scale system.
Next, we demonstrated a two-qubit logic operation via the exchange coupling of two electron spins. By combining a switchable exchange interaction using electrical pulses with single qubit rotations, we realize controlled-NOT (CNOT) operations and measure clear anti-correlations in the two-spin probabilities of the CNOT gate by independently reading out both qubits.
We then explore qubits composed of either one or three electrons, with the excess electron spin residing in the lower or upper conduction-band valley states, respectively. We observe an opposite g-factor dependence as a function of the applied dc electric field in the two type of qubits, and subsequently, we formulate a modified effective mass theory that shows direct correlations between the spin-orbit coupling parameters and the g-factors in the two valleys.
Finally we investigate spin qubits defined in a double quantum dot which has a small g-factor difference with respect to their exchange interaction. This enables us to simultaneously drive transitions of two individual spins between the T- and T+ triplet states using an ESR frequency that is not directly resonating with any of the two spins in the system.