Thomas, Edward J; (2024) Automation and Scaling of Fast Charge Readout in Silicon Quantum Dots. Doctoral thesis (Ph.D), UCL (University College London).

Text Thomas_10194279_thesis.pdf Access restricted to UCL open access staff until 1 February 2025. Download (40MB)

Abstract

Silicon quantum dot spin qubits are a promising platform for quantum computing, owing to their long coherence times, high-fidelity and fast gate operations, and compatibility with mature CMOS industrial fabrication processes. For the most valuable applications, a quantum processor must achieve fault-tolerance, requiring integration of millions of qubits with low variability together with high-fidelity control and readout. In silicon quantum dots, radiofrequency reflectometry accelerates charge readout when the qubit is coupled to a resonant circuit, but scaling of this technique to very large qubit numbers is limited by the availability of high-frequency lines in a dilution refrigerator. In this thesis, we use classical electronics at cryogenic temperatures to multiplex a single resonator and readout line for sharing by many qubits. First, a bespoke one-to-many switch is characterised at cryogenic temperatures and evaluated for reconfigurable fast sensing of quantum dot qubits fabricated on a separate chip. The charge sensing performance is evaluated in detail to explore the limits of high-speed parallel addressing. The platform is used for rapid measurements of two quantum dot devices fabricated on a 300 mm wafer process. Next, an automated protocol is developed for characterisation of 1024 single electron transistors in a 32 by 32 array in an integrated circuit incorporating addressing electronics. Automated analysis based on image recognition detects and fits a signature of Coulomb blockade to extract key figures of merit, enabling statistical comparisons between designs and evaluation of the fabrication yield. This is first demonstrated using the established dc transport technique for charge sensing. Finally, the 1-to-1024 multiplexer is embedded in a resonant circuit to make equivalent measurements of the devices using reflectometry in under 10 minutes. The minimum integration time for a signal-to-noise ratio of 1 is found to be 527 ps.

Download activity - last month

Export as JSONXMLCSV

Download activity - last 12 months

Export as XMLJSONCSV

Downloads by country - last 12 months

Export as JSONXMLCSV

Archive Staff Only

View Item