MITIGATION OF MEMORY EFFECTS IN BETA SCINTILLATION CELLS FOR RADIOACTIVE GAS DETECTION

Detection of radioactive noble gases can provide definitive evidence of a nuclear explosion at large stand-off distances. The Automated Radioxenon Sampler/Analyzer (ARSA) developed at Pacific Northwest National Laboratory (PNNL) measures the relative concentrations of xenon isotopes using a β-γ coincidence system. Typically, a set of plastic scintillating cells are surrounded by a NaI(Tl) scintillator. The cells are evacuated and the background count rate is measured before the cell is filled with sampled air. Any radioxenon present in the air emits β particles, which are detected in the plastic cell, as well as coincident γ rays, which are detected in the NaI(Tl). When a sample count is finished, the cell is again evacuated, and another background count is taken before measuring the next air sample. Previous tests of the ARSA system have shown that latent radioactivity remains in the plastic cells after evacuation of the gases, leading to a "memory effect" in which the background count rate is dependent on the sample history. The increased background results in lower detection sensitivity. Two possible solutions to the memory effect are explored in this work: depositing a thin layer of metal on the plastic cell ("metallization"), and using an inorganic scintillating cell composed of yttrium aluminum perovskite (YAP). In both cases, the presence of inorganic material at the surface is intended to inhibit the diffusion of gases into the cell walls. In the metallization experiments, several different metals (Al, Cu, chrome) were deposited on plastic cells using electron beam lithography. The light collection performance of the cells was evaluated using standard sealed γ-ray sources and compared to a bare plastic cell. The aluminized cell demonstrated comparable light collection performance and good adhesion to the plastic and was chosen for further studies. The aluminized cell, YAP cell, and bare plastic cell were each placed in the void of a CsI(Na) well counter and injected with radioactive xenon and radon. β-γ coincidence measurements were taken before, during, and after injection of radioactive gases. The YAP cell demonstrated little or no observable memory effect, while the aluminum-coated plastic cell showed reduced latent radioactivity relative to the bare cell as expected. Although the memory effect results for the YAP cell are promising, the wall thickness is too large for the escape of the xenon x- rays into the gamma-ray detector, which is required for radioxenon detection. This paper discusses the measurement details and provides recommendations for further research and optimization.