Single Photon Counting Detectors for Low Light Level Imaging Applications

This dissertation presents the current state-of-the-art of semiconductor-based photon counting detector technologies. HgCdTe linear-mode avalanche photodiodes (LM-APDs), silicon Geiger-mode avalanche photodiodes (GM-APDs), and electron-multiplying CCDs (EMCCDs) are compared via their present and future performance in various astronomy applications. LM-APDs are studied in theory, based on work done at the University of Hawaii. EMCCDs are studied in theory and experimentally, with a device at NASA's Jet Propulsion Lab. The emphasis of the research is on GM-APD imaging arrays, developed at MIT Lincoln Laboratory and tested at the RIT Center for Detectors. The GM-APD research includes a theoretical analysis of SNR and various performance metrics, including dark count rate, afterpulsing, photon detection efficiency, and intrapixel sensitivity. The effects of radiation damage on the GM-APD were also characterized by introducing a cumulative dose of 50 krad(Si) via 60 MeV protons. Extensive development of Monte Carlo simulations and practical observation simulations was completed, including simulated astronomical imaging and adaptive optics wavefront sensing. Based on theoretical models and experimental testing, both the current state-of-the-art performance and projected future performance of each detector are compared for various applications. LM-APD performance is currently not competitive with other photon counting technologies, and are left out of the application-based comparisons. In the current state-of-the-art, EMCCDs in photon counting mode out-perform GM-APDs for long exposure scenarios, though GM-APDs are better for short exposure scenarios (fast readout) due to clock-induced-charge (CIC) in EMCCDs. In the long term, small improvements in GM-APD dark current will make them superior in both long and short exposure scenarios for extremely low flux. The efficiency of GM-APDs will likely always be less than EMCCDs, however, which is particularly disadvantageous for moderate to high flux rates where dark noise and CIC are insignificant noise sources. Research into decreasing the dark count rate of GM-APDs will lead to development of imaging arrays that are competitive for low light level imaging and spectroscopy applications in the near future.