Studies of transition-edge sensor physics : thermal models and noise

Abstract

This thesis focuses on studying the noise in transition-edge sensors (TES). More specifically, the aim is to find an explanation for the observed excess noise that is limiting their performance and has been troubling the TES field in recent years. Several theories have been put forth but a definitive answer is still missing. In the early stages of this thesis work a novel theory was presented for the noise in a special TES geometry. In our theory the excess noise is caused by correlated fluctuations of superconductivity at the phase boundary between normal and superconducting states. Data from more recent experiments does not give solid support for the theory and in this thesis the validity of the model is discussed. Measurements of the complex impedance of TES detectors have shown that the thermal circuit of a TES can be complicated and imply the presence of noise caused by internal thermal fluctuations. The thermal circuit can be presented by a block model and in this thesis we try to identify the simplest model that can explain all the measured impedance and noise data. The determination of the thermal model depends on the interpretation of the complex impedance data. The measurement setup has a limited bandwidth and often the TES impedance is such that it is difficult to find the correct high frequency limit of impedance, which is an important parameter. In this work a method was developed that could be used to extract the needed information from DC data only. We have shown that a three-block thermal model is often sufficient for our data. One of the blocks was identified as originating from the underlying SiN membrane that is used for thermal isolation of the TES. The main result of this thesis is the identification of the other two blocks. These are the normal and superconducting phases inside the TES. According to our model, the observed excess noise arises from the finite thermal conductance between the two phases, which means that the superconducting part will decouple thermally from the normal part at high frequencies. Our data also shows that the jump in heat capacity at the critical temperature as predicted by the BCS theory may be reduced in thin bilayer films. We discuss possible design choices that could minimize the noise due to the N-S decoupling.