氡氣即時偵測器的發展與在臺灣地震站的部署

Development of Real-Time Radon Detector and Deployment in Seismic Station in Taiwan

吳致平(Chih-Ping Wu)

指導教授 ： 裴思達

國立臺灣大學/理學院/物理學研究所/碩士(2021年)

https://doi.org/10.6342/NTU202103773

摘要

氡222是一種具有放射性的惰性氣體，該同位素來自於鈾238的衰變鏈，其常見於許多岩石中。而其可能在地震前會釋放得較多較快，氡氣的大氣濃度被認為與地震有關係。目前常用的氡氣研究方法是使用α粒子偵測器去量測，而對於γ射線的研究不多，且較為粗糙地去計算放射量，因此有必要發展較全面的偵測系統。
偵測器由一塊碘化納圓柱體、一個光電倍增管與一個含有電路與前置放大器的基底組成。偵測器吸收γ射線的能量並將訊號放大後傳給資料擷取器，而擷取器會將訊號存成數據以待分析。在部署前，偵測器需要先做校驗以確定偵測能量與輸出電壓間的關係。校驗時會使用四個標準放射源與一袋北投石，一種具放射性的岩石，用其特定的放射能來校驗。此外在Geant4上會進行偵測器在氡氣環境中的虛擬實驗來模擬偵測器的工作情況，而結果將會用於之後的分析上。
γ偵測器會部署在花蓮氣象站的地下室中運作。除了硬體的架設外，將採用數種γ射線能譜的分析方法去評估環境中的氡氣量。第一種方法是方程式擬合，使用一個指數項與一個高斯項去分離訊號與背景雜訊。另外一種是計算在特定譜窗內的所有事件。第三種是模板擬合，運用模擬結果與在實驗室取得的背景雜訊去對部署時的能譜做擬合。之後討論關於系統與各分析方法的穩定性與敏銳度。此外，前兩種分析方法會著重在0.295、0.609與1.460百萬電子伏特的光峰。其中帶有0.295與0.609百萬電子伏特的光子來自於氡氣的衰變，而1.460百萬電子伏特的光子來自於摻在偵測器的鉀元素。
目前的分析結果顯示使用方程式擬合的估計在三種方法中是最為穩定的，其他兩種方法都造成很大的波動，因此目前最好的選擇是以方程式擬合，由0.295與0.609百萬電子伏特的所有訊號量的加總。根據該估計，穩定性保持在約7.5%。目前地震次數不足以和分析結果比較，已有的資料也未顯現其彼此的相關性。
在部署期間有發現幾個問題，一個是過高雜訊的干擾，另一個是能譜會隨時間偏移與伸縮。這些因素會影響能譜的分析結果，所以改進工作會放在以雜訊更少的偵測器取代，以及結合多個偵測器來做重合，以減少雜訊，提高訊雜比。而在部署時能譜的偏移與伸縮可以設法在分析時，嘗試以已標定的譜線來做修正。

關鍵字

伽馬射線 ； 光電倍增管 ； 閃爍體 ； 碘化納 ； 能譜分析

並列摘要

Rn-222 is a kind of noble gas with lots of radiation. The isotope is from U-238 decay chain usually found in rocks, and its concentration in atmosphere is considered with relation to earthquake because radon may be released more and faster before earthquake happens. So far, the common way for radon research used alpha ray detector to measure radon. The studies on gamma ray was few and just roughly counted amount of rays. Thus, it is significant to develop a system for gamma ray detection.
A detector consisted of a sodium iodide cylinder, a PMT and a base with a circuit and a preamplifier. The detector absorbed the energy of gamma ray and amplified signal for DAQ. The DAQ stored the signal as data for analysis. Before deployment, the detector needed calibration to determine the correlation between energy and output voltage. Four standard radioactive sources and a bag of hokutolite, a kind of rocks with lots of radiation, provided gamma ray with specific energy for calibration. Besides a simulation on Geant4 was built for simulating the performance of the detector under the radiation from radon. The result of simulation would help for the analysis.
A set of gamma ray detector would be deployed and running in a basement under a weather station in Hualien. In addition to hardware building, several analysis methods of gamma ray spectrum were utilized for evaluating radon in environment. First method was function fitting, using an exponential and a Gaussian term to separate signal and background noise. Another one was counting all events in a spectrum window covering a focused peak. The third one was template fitting, using the result of the simulation and a background spectrum in laboratory to fit the spectrum during deployment. Then there is a discussion about stability and sensitivity of the system and analysis methods. By the way, the first two method would focus on the peak of 0.295, 0.609 and 1.460 MeV. The photon energy of 0.295 and 0.609 MeV was from radon decay, and the energy of 1.460 MeV was from potassium blending into the detector.
The result of analysis for now was the evaluation by function fitting was the stablest one in three analysis methods. There were large waving for evaluation by other two methods. Therefore, the best choice so far was the evaluation by function fitting, and the evaluation was the summation of signal events from the peaks of 0.295 and 0.609 MeV. According to this evaluation, the stability kept about 7.3%. So far, the times of earthquake were not enough to study with the analysis result, and the correlation between them was not found yet with the collected data.
There were some problems were found during deployment. One was the interference from high noise, and another one was shifting and stretching by time in the spectrum. These factors influenced the result of spectrum analysis. Therefore, the improvement focuses on replacing with the detector with lower noise, and combining multiple detectors for coincidence, to reduce the noise and enhance signal to noise ratio. The correction in spectrum analysis by determining the known peaks is tried to correct the shift and stretch during deployment.

並列關鍵字

Gamma ray ； Photomultiplier ； Scintillator ； Sodium iodide ； Spectrum analysis

參考文獻

Ching­-Chou Fu, Lou­-Chuang Lee, Tsanyao Frank Yang, Cheng-­Horng Lin, Cheng-Hong Chen, Vivek Walia, Tsung-­Kwei Liu, Dimitar Ouzounov, Gioacchino Giuliani, Tzu­-Hua Lai, et al. Gamma ray and radon anomalies in northern taiwan as a possible preearthquake indicator around the plate boundary. Geofluids, 2019, 2019.

MA Vasiliev, P Blum, G Chubarian, R Olsen, C Bennight, T Cobine, D Fackler, M Hastedt, D Houpt, Z Mateo, et al. A new natural gamma radiation measurement system for marine sediment and rock analysis. Journal of Applied Geophysics, 75(3):455–463, 2011.

Vanessa Chisté, Marie-­Martine Bé, and Christophe Dulieu. Evaluation of decay data of radium­226 and its daughters. In International Conference on Nuclear Data for Science and Technology, pages 77–80. EDP Sciences, 2007.

B Badalamenti, G Capasso, ML Carapezza, W D’Alessandro, F Di Gangi, IS Diliberto, S Giammanco, S Gurrieri, PM Nuccio, F Parello, et al. Soil gas investigations during the 1991–1993 etna eruption. Acta Vulcanol, 4:135–141, 1994.

S Giammanco and M Valenza. Soil gas radon activity on mt. etna. ACTA VULCANOLOGICA, 8:222–222, 1996.