Overcoming the Low Energy Excess to Advance Sub-GeV Rare Event Detection

Transition-Edge Sensors (TESs) are a key technology in rare-event searches, including low-mass dark matter detection, but their performance at low thresholds is limited by the persistent Low Energy Excess (LEE). This work investigates whether the LEE elevates the intrinsic noise of the TES devices. Using a combination of measured noise traces and semi-analytic power spectral density (PSD) models generated via the qetpy framework, I perform a detailed noise analysis across superconducting and normal bias regimes. The close agreement between measured and modeled PSDs, particularly in the low-frequency regime, provides strong evidence that the LEE does not create an excess shot noise source in the TES. Additional comparisons between devices fabricated on Si and SiC substrates reveal substrate-dependent noise features consistent with variations in phonon transport and readout electronics, but no indication of LEE-related excess noise. This thesis finds that the LEE does not have a significant impact on intrinsic TES noise performance and highlights the utility of noise spectroscopy as a diagnostic for detector optimization.