High precision readout of superconducting resonators: For analysis of slow noise processes

10 years ago superconducting resonators emerged with two potential applications, firstly kinetic inductance detectors were proposed and secondly the strong coupling of a resonator to a qubit was realized. As such, the field of superconducting quantum circuits is beginning to enter maturity and has become competitive with alternative approaches to quantum information processing and photon detection. However, there remains much room for further improvement and some problems persist still. Prevalent among these problems is a high level of environmental noise interfering with the device. This environment can be parametrised as a bath of two level fluctuators (TLFs): an uncontrolled intrinsic two level system which couples to the superconducting quantum circuit. These can produce the ubiquitous 1/f noise and are a dominant source of decoherence.

In this thesis, microwave resonators are interrogated using a novel high-precision frequency readout technique. This technique is based upon Pound locking and is implemented to examine the effects of two level fluctuators. The Method uses feedback to track deviations in the centre frequency of a microwave resonator. Sensitivity to Hz-level fluctuations allows the technique to trace ultra low dielectric loss tangents. Hence the approach is suitable for accessing even the samples with a very low density of two level fluctuators. Additionally a feedback mechanism allows for slow noise processes to be studied with exceptional statistical confidence.

This thesis outlines the theory and development of the Pound loop. This is characterized and improved until capable of resolving Hz-level deviations and operating towards single photon energies within the resonator. This allows the 1/f noise of a resonator to be studied under varying microwave drive and temperature. Ultimately we show that a new design of epitaxially grown superconductor is able to realize an ultra low dielectric loss tangent, and a low noise level, probably significantly below that measured by any other system. Which allows the temperature dependence of 1/f noise in superconducting resonators to be examined. This result provides important new guidance to methods that might eradicate the TLF problem from superconducting quantum circuits.