Advisor: Prof. Stephen A. Lyon, Princeton University
Peter Mark Award for "Excellence in Material Science Research", Princeton School of Engineering and Applied Sciences (SEAS).
My senior thesis project focused on the first implementation of quantum dots for electrons-on-helium, to enable ensemble quantum computing and simulation. Since single-electron detection is challenging for electrons-on-helium, our device intended to improve read-out fidelity by simultaneously measuring several quantum dots. I developed nano-fabrication techniques to construct large arrays (~ 10^9) of nearly identical metal-oxide quantum dots. This included extreme fine-tuning of the photolithography and etching steps to ensure uniform shape and depth of the individual dots across the array. I also worked on exact and approximate (Hartree Fock and Constant-Interaction) theoretical models to predict the energy transitions within these many-electron quantum dots.
The size of each dot within these arrays can be varied as per the scientific application. Larger dots (>1um) can be used to simulate atoms by populating them with many electrons, whereas smaller ones (< 100nm) can be used for quantum computing through electron-spin-resonance (ESR) measurements.
Senior ThesisAdvisor: Prof. Alexandre Blais, Université de Sherbrooke
We explored the 'quantum demolition' effects of dispersive readout in circuit QED. Theoretical models predict that qubit relaxation increases with the mean photon number in the associated readout resonator. We compare these models with experimental data from qubits qubits available through the IBM Quantum Experience (IBM Qiskit Pulse) as well through collaboration with experimental groups. Since the research is on-going, some experimental results have been witheld from the final report.
Final Report I: Theory Final Report II: Experiments with QiskitAdvisor: Prof. Andrew Houck, Princeton University
We develop a set of nano-fabrication techniques which consistently achieve increased T1 coherence times for Transmon qubits - with an average exceeding 0.30ms, almost three-times the longest T1 previously published (0.114ms). Our recipes eliminating sources of dielectric loss in two ways: