Graduate Thesis Or Dissertation | A Beta-Gamma Radioxenon Detection System using Ultra-Bright Inorganic Scintillators and Solid State Detectors | ID: fn1075224 | translation missing: pt-BR.hyrax.product_name
The atmospheric detection of four radioxenon isotopes (131mXe, 133mXe, 133Xe, and 135Xe) released during a nuclear detonation is a key tool utilized by the Comprehensive Nuclear-Test-Ban Treaty Organization (CTBTO) to identify clandestine nuclear weapon testing activity. These radioxenon isotopes all decay via the near-simultaneous release of an electron and a photon, which allows them to be easily discriminated from background at extremely low concentrations (≤ 1 mBq/m3). Detection systems employed in the International Monitoring System (IMS) that utilize this technique, though effective, make use of costly and archaic technologies which yield suboptimal energy resolutions and are subject to memory effect.
The PIPS-SrI2(Eu) is a prototype radioxenon detection system that makes use of modern technologies to address these problems and work towards improving performance. This system utilizes custom D-shaped SrI¬2(Eu) scintillators coupled to silicon photomultipliers as photon detectors and a pair of passivated implanted planar silicon wafers for electron detection. Coincidences are identified in hardware in real-time using a field programmable gate array-based multi-channel digital pulse processor. The system demonstrates a memory effect of 0.318 ± 0.026%, a ~15× improvement relative to plastic scintillators. Minimum Detectable Concentration (MDC) estimates in terms of mBq/m3 air calculated for 131mXe, 133mXe, 133Xe, and 135Xe, respectively, using the parameters from the Xenon International gas processing unit and assuming a black sample and zero memory effect yield sensitivities of 0.12 ± 0.03, 0.27 ± 0.05, 0.15 ± 0.02, and 1.00 ± 0.08, respectively. These optimistic MDC estimates compare well with other radioxenon detection systems utilized in the International Monitoring System (IMS). A new spectral deconvolution approach using a maximum likelihood and region sectioning, designated Regional Spectral Deconvolution (RSD), was also designed and tested. This method increased convergence speed by nearly two orders of magnitude and performed similarly to traditional spectral deconvolution methods in terms of accuracy. However, RSD did not demonstrate significant improvement in low count rate situations that was expected when compared to these traditional spectral deconvolution methods.