A stilbene - CdZnTe based radioxenon detection system

A stilbene - CdZnTe based radioxenon detection system

Journal of Environmental Radioactivity 204 (2019) 117–124 Contents lists available at ScienceDirect Journal of Environmental Radioactivity journal h...

2MB Sizes 0 Downloads 2 Views

Journal of Environmental Radioactivity 204 (2019) 117–124

Contents lists available at ScienceDirect

Journal of Environmental Radioactivity journal homepage: www.elsevier.com/locate/jenvrad

A stilbene - CdZnTe based radioxenon detection system Harish R. Gadey , Abi T. Farsoni, Steven A. Czyz, Kacey D. McGee


School of Nuclear Science and Engineering, Oregon State University, 3451 SW Jefferson Way, Corvallis, OR, 97331, USA



Keywords: CTBTO Memory effect Nuclear weapon explosion monitoring Radioxenon detection Stilbene

Atmospheric monitoring of radioxenon is one of the most widely used methods by the Comprehensive NuclearTest-Ban Treaty Organization (CTBTO) to detect elevated levels of 131mXe, 133/133mXe, and 135Xe. The ratios of these radionuclides help discriminate between peaceful use of nuclear technology and nuclear weapon explosions. Radioxenon detection systems often use plastic scintillators in the capacity of an electron detector and a gas cell, plastic gas cells are responsible for introducing high memory effect in these systems. This work presents the design of a new detection system for radioxenon monitoring that utilizes silicon photomultipliers, a stilbene gas cell, and a CdZnTe detector. This detector was evaluated using xenon radioisotope samples produced in the TRIGA reactor at Oregon State University. A 48-h background was collected and calculations of the Minimum Detectable Concentration (MDC) were carried out using the Region of Interest (ROI) approach. An MDC of less than 1 mBq/m3 was obtained for 131mXe, 133Xe, and 133mXe in accordance with the sensitivity limits set by the CTBTO and performs respectably when compared to state-of-the-art radioxenon detection systems. Using 131mXe, this study indicates that the stilbene gas cell exhibits a memory effect of 0.045 ± 0.017%, this is almost a twoorder magnitude improvement compared to plastic scintillators. The primary purpose of this work is to explore the use of new stilbene detection media for radioxenon application and addressing the problem of memory effect.

1. Introduction The Comprehensive-Nuclear-Test ban Treaty (CTBT) prohibits the testing of nuclear weapons on the face of the earth; this includes atmospheric, underground, surface, and underwater testing (Saey and De Geer, 2005; CTBTO, 2019). This treaty was signed and ratified by 184 and 167 nations respectively (CTBTO Map, 2019). In an effort to provide member nations with a mechanism for monitoring and verifying that other states are fulfilling their treaty obligations, the CTBTO formed the International Monitoring System (IMS). The IMS consists of a series of 321 monitoring stations and 16 national labs worldwide that look for signs of nuclear weapon explosions (CTBTO Map, 2019). These stations take advantage of four different monitoring technologies: infrasound, hydroacoustic, seismic, and radionuclide to help scientists classify if an event is nuclear in nature. The CTBTO plans to install a total of 80 radionuclide monitoring stations all over the world out of which 40 will be additionally equipped with noble gas detection (Zähringer et al., 2009). A nuclear weapon explosion on the surface of the earth, underwater, or in the atmosphere, would be relatively easy for satellites to identify. Underground explosions, on the other hand, are not readily observed by satellites which is why the scientific community relies on other means ∗

(Carrigan and Sun, 2014). In the event of an underground nuclear explosion, large amounts of fission products are confined in the blast chamber making it hard to readily identify them. Observing the fission yield curve it can be seen that noble gases like xenon have a large independent and cumulative yield (Bowyer et al., 2002). This works to our advantage because xenon, being a noble gas, is chemically inert and can escape the blast chamber with relative ease compared to other fission fragments (Carrigan et al., 1996). The xenon isotopes of interest for nuclear weapon explosions are 131mXe, 133Xe, 133mXe, and 135Xe. It must be noted that normal reactor operations also produce some of these radionuclides; nevertheless the ratios between these xenon isotopes is quite different for reactor operations and a nuclear weapon tests, which is what allows for the distinction between normal reactor operations and a nuclear explosion (Kalinowski and Tuma, 2009; Le Petit et al., 2008). The half-lives, decay energies, and branching ratios of these four xenon isotopes are provided in Table 1. The noble gas detection systems in the IMS use gas filtration units to collect atmospheric gas samples and separate xenon chemically before it is injected into the detectors. The half-lives of all the aforementioned xenon isotopes are between 9.2 h (0.38 days) and 11.93 days; this timescale is of particular interest because it is neither too short for the radionuclides to decay before being detected and neither too long for them to saturate

Corresponding author. 3451 SW Jefferson Way, Corvallis, OR, 97331, USA. E-mail address: [email protected] (H.R. Gadey).

https://doi.org/10.1016/j.jenvrad.2019.03.027 Received 29 January 2019; Received in revised form 27 March 2019; Accepted 27 March 2019 0265-931X/ © 2019 Elsevier Ltd. All rights reserved.

Journal of Environmental Radioactivity 204 (2019) 117–124

H.R. Gadey, et al.

Table 1 Coincident decay properties of radioxenon isotopes of interest (Czyz et al., 2019).

the atmosphere. Also, of particular interest is the decay scheme of these radionuclides; all the four xenon isotopes of interest decay via electronphoton coincidence decay. This means when any of these radionuclides decay, they release an electron in the form of a beta particle or conversion electron and a photon in the form of gamma or x-ray virtually simultaneously. This significantly helps in rejecting background events which would otherwise trigger the system thereby making the detector sensitive to low amounts of radioxenon. During the preparatory committee meetings of the CTBTO, four countries (USA, Russia, Sweden, and France) expressed interest in developing detectors for monitoring atmospheric radioxenon. The American Automated Radioxenon Sampler/Analyzer (ARSA) developed by Pacific Northwest National Lab (PNNL) consists of a plastic gas cell, which doubles up as the electron detector, and a well-type NaI detector for photon detection (Cooper et al., 2007). Both detectors operate in coincidence mode, generating 2-D spectra over which regions of interests (ROIs) are identified for various xenon isotopes. The Swedish Automatic Unit for Noble gas Acquisition (SAUNA) designed and

developed by the Swedish Defense Research Agency (FOI) is also an electron-photon coincidence system. It consists of a plastic gas cell and a NaI detector working in coincidence (Ringbom et al., 2003). The initial version of the Russian Analyzer for Xenon Radioisotopes (ARIX) developed by the Khlopin Radium Institute was an electron gated gamma spectrometer which had a thin layer of polystyrene deposited on the NaI detector. The later developed version of ARIX was based on electron-photon coincidence (Popov et al., 2005; Prelovskii et al., 2007). The French Système de Prélèvement Automatique en Ligne Avec l’Analyse du Xénon (SPALAX™) developed by the French Atomic Energy Commission (CEA) functions on high-resolution gamma spectroscopy (Le Petit et al., 2015, 2008). Recently Iran has also developed a Noble Gas Analysis System for radioxenon measurements based on coincidence detection (Doost-Mohammadi et al., 2016). Some of these systems have been successful in detecting the North Korea weapon tests and emissions from the Fukushima plant at various places all over the world (Becker et al., 2010; Bowyer et al., 2011, 2011; Orr et al., 2013; Ringbom et al., 2014; Saey et al., 2007; Shilian et al., 2013). Although 118

Journal of Environmental Radioactivity 204 (2019) 117–124

H.R. Gadey, et al.

these systems have high electron-photon coincidence detection efficiency (ARSA, SAUNA) and improved spectral resolution (SPALAX™) they suffer from downsides that include complex gain matching and bulkiness. Particularly detrimental to systems that use plastic scintillators is memory effect; the phenomenon where residual xenon that had diffused in the plastic scintillator from a previous measurement interferes in subsequent measurements. This primarily leads to elevated backgrounds and causes a problem when performing back-to-back measurements is desirable. A few other designs studied to remedy some of the short-comings for radioxenon detection include Phoswich and PIPS-CZT systems (Alemayehu et al., 2014; Czyz et al., 2019; Ely et al., 2005; Farsoni et al., 2013; Hennig et al., 2007; Sivels et al., 2017). To address some of these issues, research at Oregon State University has focused on developing compact low-cost radioxenon detection systems. The Two Element CZT (TECZT) detector developed by (Ranjbar et al., 2017) uses two face-to-face co-planar CZT detectors held by a plastic holder (Ranjbar et al., 2016). CZTs are a high-atomicnumber (Z) room temperature semiconductor detectors with excellent energy resolution. This system was successful in achieving high resolution for the photon peaks but suffered from electron backscatter because of the high atomic number elements in the CZT. To remedy this, a prototype detector using a CZT crystal, an Array of Silicon Photomultipliers (SiPMs), and a Plastic scintillator (CASP) was designed by (Czyz and Farsoni, 2017; Czyz et al., 2018). This combination of detector material and configuration was successful in solving the problem of electron backscatter but reintroduced memory effect in the plastic scintillator gas cell. Previously, efforts were undertaken to solve memory effect by using Al2O3 atomic layer deposition on the inner surface of the gas cell (Bläckberg et al., 2013, 2011; Warburton et al., 2013). One alternative approach is to entirely replace the plastic scintillator with a material that is similar in radiation detection properties, but inherently resistant to memory effect. Literature suggests that the crystalline structure of stilbene can help reduce the diffusion of xenon through the material and this, in turn, helps in the reduction of memory effect (Sivels et al., 2019; CVT, 2017; CTBTO, 2015; CVT, 2018). Stilbene is also known to offer improved energy resolution for conversion electrons. Improved energy resolution enables us to fit a tighter ROIs around the peak which helps improve the Minimum Detectable Concentration (MDC). To address some problems in the current state-of-theart and to take advantage of the most modern materials and technologies, this research work proposes the use a SiPM coupled well-type stilbene gas cell in conjunction with a coplanar CZT for radioxenon detection. The primary goal of this work is to study the response of the stilbene-CZT detection system to the four radioxenon isotopes of interest and evaluate the memory effect of the stilbene gas cell. The experimental setup, detector response to xenon radioisotopes, memory effect measurements and MDC calculations are detailed in the work.

Fig. 1. The assembled Stilbene-CZT radioxenon detection system secured inside the holder.

2.1. Coplanar CZT A coplanar CZT design was used as a detection medium for the gamma and x-ray photons. The CZT is positioned such that the cathode side is facing the stilbene gas cell and the anode side of the CZT is facing outwards. A coplanar pattern was chosen to achieve a single charge carrier design which would make the signal independent of the depth of interaction while maintaining high resolution at room temperature (Luke, 1994). This CZT detector was manufactured by Redlen Technologies and measures 10 x 10 × 10 mm3. Thin gold wires were attached to the collecting and non-collecting grid pads to enable signal readout. A biasing voltage of −1000V for the cathode and a −80V for the non-collecting grid was used as recommended by the manufacturers. 2.2. Stilbene gas cell The gas cell for this detector was fabricated using stilbene. This cell also acts in the capacity of an electron detection medium. Stilbene was chosen in favor of EJ-212 plastic scintillator (previously used in CASP) because of its improved electron resolution (CVT, 2017) and little to no memory effect (CVT, 2018). The stilbene gas cell, manufactured by Inrad Optics, is made by attaching two end caps on the top and bottom of the body. All three parts of the gas cell are made from the same material. An outward facing barb is attached to the top end cap of the gas cell which is connected to a plastic tube. This plastic tube has a 3way Luer lock attached on the other side to enable vacuum generation and gas injection. The well-type detector has an external diameter of 18.7 mm, a total height of 21.7 mm and a wall thickness of 1.8 mm. MCNP simulations were carried out to confirm that the 1.8 mm thickness of stilbene wouldn't attenuate the 30 keV x-ray significantly (an attenuation of ∼5.4% was observed). The end caps have a thickness of 2 mm. This geometry provides a total volume of 3.17 cm3 and a solid angle of approximately 3.9 π. The underside of the stilbene gas cell is coupled to the PCB mounted SiPM array using optical grease. In an effort to further improve light collection the sides of the gas cell were wrapped in several layers of Teflon.

2. Detector design The prototype system consists of two detector elements; a coplanar CZT detector and a stilbene scintillator coupled to an array of 3 x 3 SiPMs. The two detection elements are held in a holder made by a 3D printer. The top of the gas cell hosts a plastic barb which connects the gas cell and the injection tube. The bottom surface of the stilbene gas cell is coupled to the SiPMs using a silicon optical grease. The current pulse from the SiPMs is then routed through a pre-amplifier. The signal from the collecting and non-collecting electrodes of the CZT is routed through the pre-amplifiers followed by a subtraction circuit. A 2channel digital pulse processor (DPP2) is used to read the data from the CZT and stilbene + SiPM. The pulse processor houses a Field Programmable Gate Array (FPGA) which is used to identify coincidence pulses in real time. These pulses are then sent to the PC to be processed by MATLAB. The assembled detector is shown in Fig. 1.

2.3. SiPM array Silicon photomultipliers (SiPMs) in their inherent form are arrays of thousands of very small reverse biased photodiodes operating in the Geiger mode. When a diode experiences an interaction with a photon, an avalanche of electrons occurs, creating a small fixed current with high gain. When coupled to a scintillator, the combined signal from 119

Journal of Environmental Radioactivity 204 (2019) 117–124

H.R. Gadey, et al.

thousands of these small diodes triggered by visible light photons provide information about the energy of impinging radiation. SiPMs were chosen as opposed to conventional photo-multiplier tubes (PMTs) primarily to reduce the size of the detection system, decrease voltage requirements, improve ruggedness, and to achieve a light collection mechanism resistant to magnetic fields. These SiPMs can also be mounted directly on a PCB which can then be coupled to the Stilbene gas cell. For this work, a 3 x 3 SiPM array was used consisting of SensL J-series 6 × 6 mm2 P-on-N (MicroFJ-60035-TSV-A1) cells. The J-Series SiPMs have a quantum efficiency of greater than 50% at 420 nm. The SiPM array has a dimension of 18.79 × 18.79 mm2 and is subjected to a voltage of −26.5 V (SensL, 2019).

3. Experimental work 3.1. Detector calibration and sample preparation Black masking tape was used to wrap the outward facing barb to ensure complete light sealing and to prevent external photons from interacting with the SiPMs. A 10 μCi 137Cs lab check source was used to first calibrate the CZT. The threshold for the CZT and stilbene was found to be about 20 and 30 keV, respectively. The cathode and the grid were subjected to the voltages recommended by the manufacturer and the CZT was exposed from the cathode side to prevent interactions from happening too close to the anode which would have resulted in variable weighting potentials between the electrodes resulting in degraded energy resolution (He et al., 1997, 1996). Spectra were collected for various trimmer settings to obtain the best energy resolution. A 2.3% FWHM for the 662 keV peak was achieved when the CZT detector was exposed to a 137Cs source. Once the CZT was calibrated, the stilbeneSiPM was calibrated using the Compton backscatter line from the 662 keV of 137Cs source. The stilbene/CZT spectra and 2-D coincidence plot, including this Compton backscatter line, are shown in Fig. 2a and Fig. 2b respectively. While carrying out the backscatter coincidence measurements a large amount of noise was recorded by the CZT at low energies (< 20 keV) possibly from ground loops or external noise. Therefore, in an effort to obtain a clean spectrum counts in channels below 22 keV from the CZT were rejected for this measurement. After the detectors were calibrated, xenon samples were prepared by loading polystyrene syringes with pure samples of commercially available stable and enriched xenon isotopes (Ranjbar et al., 2017). These syringes are then loaded in the thermal column of the Oregon State Universities TRIGA reactor. Each sample loaded in the syringes were between 0.5 and 1.0 mL at atmospheric pressure depending on the pressure the gases are maintained at; the syringes were then exposed to a thermal neutron flux of approximately 7 × 1010 n cm−2 sec−1. A total of 5 irradiations were carried out — two 134Xe, one 132Xe and, two 130 Xe. While measuring 135Xe for the first-time triggering issues were experienced which was later resolved during the second injection. Xenon131m was injected twice; once to measure the detector response and carry out memory effect measurements and the second time to validate memory effect results which were initially obtained. Xenon133/133m was injected once to measure the detector response.

2.4. Readout electronics A PCB was designed in-house to process the signals from the collecting and non-collecting grids of the CZT and the signal from the stilbene and SiPM. This is the same board used for the CASP system. The board was designed to enable direct mounting of the SiPMs, stilbene gas cell and the coplanar CZT. The preamplifier circuit for the SiPMs was designed as recommended by the manufacturer. The signals from the collecting and non-collecting grids of the CZT were routed through two Amptek A250F/NF pre-amplifiers. The output signals from the preamplifiers are then processed by a low power voltage feedback amplifier AD8039. This amplifier performs the signal subtraction of the noncollecting grid from the collecting grid to produce a depth independent signal from the CZT. A trimmer (variable resistance) was used to perform gain adjustments in the interest of obtaining the best x-ray/ gamma-ray energy resolution. Following this conditioning, the analog signals from the CZT and SiPM are then read by a 2-channel digital pulse processor designed at Oregon State University. The DPP2 houses two 12-bit, 200 MHz ADCs and a Xilinx Spartan-3 FPGA. For this work, the FPGA firmware is programmed to identify coincident events from radioxenon decay in real-time. Parameters like the trigger thresholds, gain, and coincidence time window (CTW) can all be set via the PC using the MATLAB user interface. If the system receives a signal from both the CZT and Stilbene above the set threshold within a user-defined coincidence time, the signal data is transferred via USB to the MATLAB user interphase for processing. For the radioxenon measurements, the coincidence time window was set at 0.625 μsec. A trapezoidal filter was realized in MATLAB with a peaking time of 1 μsec and flat top time of 0.9 μsec for processing CZT pulses and 0.625 μsec peaking time and 1 μsec flat-top time for processing pulses from the SiPM.

3.2. Xe131m Xe131m emits a 30 keV x-ray and releases a 129 keV, 159 keV conversion electron. Out of these three emissions, the 30 keV x-ray and the 129 keV conversion electron are released in coincidence. This xenon isotope is important among radioxenon because it is the only one which

Fig. 2. a) CZT and stilbene spectra from 137Cs backscatter coincidence; b) 2D coincidence plot from CZT and Stilbene. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) 120

Journal of Environmental Radioactivity 204 (2019) 117–124

H.R. Gadey, et al.

Fig. 3. Xe131m coincidence measurements; a) CZT and Stilbene 1-D plots; b) 2-D decay plots from the detection media. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 4. Xe133/133m coincidence measurements; a) CZT and Stilbene 1-D plots; b) 2-D decay plots from the detection media. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 5. Photon gated electron spectra from


Xe decay showing; a) the 45 and 199 keV conversion electron; b) 346 keV (βmax) continuum.

has a monoenergetic electron emission. This helps in carrying out accurate detector calibration and characterization. Fig. 3a shows the CZT and stilbene spectra collected from 131mXe. The 30 keV x-ray and the 129 keV conversion electron resolution was calculated to be 25.8% and 20.2% respectively. In the 2D coincidence plot, the electron energies deposited in the stilbene are shown on the x-axis and the photon energies sensed by the CZT are shown on the y-axis. Since both the emissions are monoenergetic in nature we expect to see a confined area where counts are populated (Fig. 3b).

coincidence with a 30 keV x-ray photon. Xe133 decay has two branches, an 81 keV gamma photon in coincidence with a 346 keV (βmax) electron and a 30 keV x-ray in coincidence with a 346 keV beta and a 45 keV conversion electron. The CZT should be sensing two peaks corresponding to the release of an x-ray and a gamma photon, both these photons were detected and a resolution of 9.3% FWHM was achieved for the 81 keV peak (Fig. 4a). The SiPMs would detect a beta spectrum ranging from 0 to 346 keV in coincidence with the 81 keV gamma photon (first branch of the 133Xe decay), this feature is shown in the 2-D spectra (Fig. 4b). The SiPMs would also be seeing a 199 keV conversion electron in coincidence with the 30 keV x-ray (133mXe) and a beta spectrum (0–346 keV) in coincidence with a 45 keV conversion electron and a 30 keV x-ray (second branch of the 133Xe decay). All three features from the decay are shown in the 2-D spectra. Fig. 5a and Fig. 5b

3.3. Xe133/133m Upon irradiation of 132Xe , two xenon isotopes are produced; 133mXe and 133Xe. Xe133m releases a 199 keV conversion electron in 121

Journal of Environmental Radioactivity 204 (2019) 117–124

H.R. Gadey, et al.

Fig. 6. Xe135 coincidence measurements; a) CZT and Stilbene 1-D plots; b) 2-D decay plots from the detection media. For interpretation of color please refer to the online version of the manuscript. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

presents the x-ray (31 keV) and gamma (81 keV) gated electron spectra highlighting individual contributions from the 45 and 199 keV conversion electron.

was then calculated by taking the ratio of the count rate after each pump and the baseline count rate. After five pumps a memory effect of 0.046 ± 0.014% and 0.044 ± 0.031% was achieved during the first and second injection respectively. Using both the experimental runs and taking into account error propagation the memory effect of the stilbene gas cell was determined to be 0.045 ± 0.017%. This is a two orders of magnitude reduction in memory effect compared to approximately 5% observed in plastic scintillators (Warburton et al., 2013).

3.4. Xe135 Xe135 has the shortest half-life (9.2 h) amongst the four radioxenon isotopes of interest. The decay of this radionuclide releases a 250 keV gamma photon in coincidence with a beta particle with energy between 0 and 910 keV. Both these features were observed while measuring 135 Xe, a total of 5 million pulses were collected; the 1-D and 2-D coincidence plots are shown in Fig. 6a and Fig. 6b respectively. In the 2-D coincidence spectra, it can be seen that the counts are populated in a narrow region around the 250 keV gamma photon and stretch from 0 to 910 keV along the stilbene axis. The resolution of the 250 keV gamma peak was measured to be 4.2% FWHM.

3.6. Minimum Detectable Concentration (MDC) MDC is a measure of the minimum amount of activity that can be measured in a sample of air with a 95% accuracy (true-positive). The MDC of a system is calculated in the units of mBq per m3 of air. For systems to be eligible for use in the International Monitoring Systems they are required to attain an MDC of less than 1 mBq/m3 for 133Xe. The MDC is calculated according to the following equation:

3.5. Memory effect


Memory effect is the phenomenon where a portion of the gas sample from one measurement diffuses through the detection media and causes interference in the subsequent experimental runs (Bläckberg et al., 2013). Stilbene, from literature review, has proven to be one of the choices to curb memory effect. For the evaluation of memory effect, two 131m Xe samples were injected in the gas cell; the first sample was used to get an initial estimate of the memory effect and the second sample was injected after a few weeks to validate the initial results. Xe131m was the choice of radionuclide because of its long 11.93 days half-life which would nullify the need to conduct decay corrections since all the measurements were carried out in the span of a few hours. A tubing system was used to connect the gas cell to the roughing pump, and a vacuum gauge is connected to this line for monitoring pressure in the gas cell. A valve is located between the gas cell and the roughing pump for the purpose of preventing back-flow once the pump is turned off. After the gas is injected in the cell, a fixed number of coincidence counts and the time required to collect these counts were recorded. A 3σ full-width ROI was established around the 129 keV conversion electron peak and the 31 keV x-ray peak. From this information, the baseline coincidence count rate was calculated by dividing the counts in the ROI by the data collection time. The roughing pump was then connected to the gas cell and suction was applied for 5 min attaining a pressure of about 5 torr in each run. After this the valve between the gas cell and the pump was closed to prevent back-flow, the luer lock on the gas cell was shut to maintain the vacuum in the cell, the tubing was disconnected, and the roughing pump was turned off. The number of counts and the time were then recorded via the DPP2 MATLAB interface. The number of counts in the ROI and the time to collect these counts help to calculate the count rate after each pump. Memory effect

mBq m3air

2.71 + 4.65








Tc ))(



TA ))

1000 Vair


Where: 0


Ts BckCnttotal + Tb + MemoryCnt +

2 BckCnt

+ InterferenceCnt +

2 InterferenceCnt

2 MemoryCnt


• : Efficiency for detection of • : Efficiency for detection of Branching ratio of • :: Branching ratio of • : Decay constant of isotope of interest [s • T : Collection time of xenon sample [s] • • T : Processing time of gas [s] • T : Counts acquisition time [s] • V : Collected air volume [m ] BR










• BckCnt : Summation of the total number of background counts observed in the ROI of interest • InterferenceCnt : Counts due to interferences from radon daughters total

and overlapping ∼30 keV regions of interest


Journal of Environmental Radioactivity 204 (2019) 117–124

H.R. Gadey, et al.

detector by radiation. In this simulation, 2 million photons/electrons of variable energies were used to estimate the efficiencies. The choice of detection media for photons and electrons was CZT and stilbene respectively. The detection efficiency was defined as the ratio of the number of interactions and the total number of particles simulated. The branching ratio for each xenon isotope of interest is provided in Table 1. Values of Tc, Tp and, TA are standard constants relevant to the gas processing system. As no gas processing system is used in this work, the values used for the ARSA detection system are utilized. The volume of air collected varies with each detection system but is dependent on the volume of the gas cells used in the system. A scaling factor is required to be applied since the active volume of the stilbene gas cell is not the same as the one used in ARSA. This factor approximately equals 0.5 times that of ARSA. Therefore, the volume of air (Vair) used in this calculation is 0.5 times that of the ARSA system. Xe133 has two decay paths, one consists of an 81 keV gamma photon in coincidence with a 0–346 keV electron and a 31 keV x-ray in coincidence with a 45 keV conversion electron and a 0–346 keV beta particle. Therefore, in such cases where there are two ROIs, a weighted average of the two decay paths is computed as shown in (4):

Fig. 7. Forty-eight-hour background coincidence 2-D plot from the StilbeneCZT detection system. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) Table 2 Comparison of MDC values of radioxenons of interest for the Stilbene-CZT, TECZT and, CASP detection systems. Radioxenon Isotope

stilbene- CZT MDC (mBq/m3)

TECZT MDC (mBq/m3)

CASP MDC (mBq/ m3)


0.483 0.381 0.836 3.336

1.45 1.85 0.51 1.47

0.586 0.169 0.838 4.321

Xe Xe Xe 135 Xe 133m 133

± ± ± ±

0.006 0.005 0.013 0.495

± ± ± ±

0.06 0.04 0.14 0.05

± ± ± ±


0.031 0.018 0.027 0.570

For this calculation, the counts from radon interference are considered to be zero because the injected samples were not collected from the atmosphere but were pure xenon samples that were irradiated in the Oregon State University TRIGA reactor thermal column. The counts from memory effect are also considered to be zero for this calculation because the stilbene gas cell has an almost negligible memory effect. Moreover, a large amount of time was allowed to pass between the injection of a sample and background measurements for the purpose of allowing any residual radioxenon to decay. Using the aforementioned assumptions, the 0 term simplifies to:


Ts BckCnttotal + Tb

2 BckCnt



81 keV 133Xe



30 keV 133Xe



For the purpose of this calculation, a 3 ROI was fit around the peak and for beta particles, the ROI extends from 0 to βmax. The number of counts that fall in this designated ROI after a 48-h measurement is considered for calculating the MDC of each radioxenon. The resolution of the 199 keV conversion electron peak from 133mXe couldn't be accurately determined, therefore, from a conservative standpoint, the resolution of this peak is assumed to be 23.3%, the same as that obtained from the plastic scintillator (EJ-212) in the CASP system. For 133m Xe, even after 48 h, there were no counts recorded in the ROI. In such a case the ROI was expanded in both the stilbene and CZT axis by a constant percentage increase of 20% per step until a count is observed. When a count is observed, the ratio of the initial and expanded ROI is used for the purpose of calculating the MDC (Cagniant et al., 2014). A total of 7 background runs were carried out for this work. In some measurements there was a large difference in the number of counts in the ROI; this can mainly be attributed to the electronic noise in the CZT. The background data used for the MDC calculations are the results from the background with the least ROI counts. Table 2 reports the MDC of the stilbene-CZT system and also includes the results obtained from other CZT based systems like the TECZT and CASP for the purpose of comparison.

• MemoryCnt : Counts due to memory effect • Ts: Sample measurement time [s] • Tb: Background measurement time [s]


mBq m3 air

3.7. Background


In this work, while recording background, over 267,000 single events were recorded by the system out of which only 400 coincidence events fall within the 1000 × 1000 keV (Fig. 7) stilbene-CZT grid, thereby giving a background rejection rate of close to 99.85%. The total and coincidence background count rate observed by the system is found to be 1.55 and 0.0023 cps respectively.

The error in the background is assumed to be a Poisson distribution in nature, under this assumption the error in the background ( BckCnt ) is simply the square root of the counts in the ROI. For the purpose of calculating the efficiency of each radiation source, the detection media were modeled in MCNP6 and an f8 tally was taken using 2E6 histories. An f8 tally provides the energy distribution of the pulses created in the

Table 3 Resolution and memory effect values of the stilbene-CZT radioxenon system. 31 keV CZT (%)

81 keV CZT (%)

250 keV CZT (%)

662 keV CZT (%)

129 keV Stilbene (%)

Memory Effect (%)






0.045 ± 0.017


Journal of Environmental Radioactivity 204 (2019) 117–124

H.R. Gadey, et al.

4. Conclusion

Czyz, S.A., Farsoni, A.T., 2017. A radioxenon detection system using CdZnTe, an array of SiPMs, and a plastic scintillator. J. Radioanal. Nucl. Chem. 313, 131–140. Czyz, S.A., Farsoni, A.T., Ranjbar, L., 2018. A prototype detection system for atmospheric monitoring of xenon radioisotopes. Nucl. Instrum. Methods Phys. Res. Sect. Accel. Spectrometers Detect. Assoc. Equip. 884, 64–69. Czyz, S.A., Alhawsawi, A.M., Farsoni, A.T., Gadey, H.R., Ranjbar, L., Mannino, M.A., McGee, K.D., March 2019. A radioxenon detection system using PIPS and CZT. J. Radioanal. Nucl. Chem. 319 (3), 703–715. https://link.springer.com/article/10. 1007/s10967-018-6367-2. Doost-Mohammadi, V., Afarideh, H., Etaati, G.R., Safari, M.J., Rouhi, H., 2016. INGAS: Iranian noble gas analyzing system for radioxenon measurement. Radiat. Phys. Chem. 120, 26–32. Ely, J.H., Aalseth, C.E., McIntyre, J.I., 2005. Novel beta-gamma coincidence measurements using phoswich detectors. J. Radioanal. Nucl. Chem. 263, 245–250. Farsoni, A.T., Alemayehu, B., Alhawsawi, A., Becker, E.M., 2013. A phoswich detector with Compton suppression capability for radioxenon measurements. IEEE Trans. Nucl. Sci. 60, 456–464. He, Z., Knoll, G.F., Wehe, D.K., Rojeski, R., Mastrangelo, C.H., Hammig, M., Barrett, C., Uritani, A., 1996. 1-D position sensitive single carrier semiconductor detectors. Nucl. Instrum. Methods Phys. Res. Sect. Accel. Spectrometers Detect. Assoc. Equip. 380, 228–231. He, Z., Knoll, G., Wehe, D., Miyamoto, J., 1997. Position-sensitive single carrier CdZnTe detectors. Nucl. Instrum. Methods Phys. Res. Sect. Accel. Spectrometers Detect. Assoc. Equip. 388, 180–185. Hennig, W., Tan, H., Fallu-Labruyere, A., Warburton, W.K., McIntyre, J.I., Gleyzer, A., 2007. A phoswich well detector for radioxenon monitoring. Nucl. Instrum. Methods Phys. Res. Sect. Accel. Spectrometers Detect. Assoc. Equip. 579, 431–436. Kalinowski, M.B., Tuma, M.P., 2009. Global radioxenon emission inventory based on nuclear power reactor reports. J. Environ. Radioact. 100, 58–70. Le Petit, G., Armand, P., Brachet, G., Taffary, T., Fontaine, J.P., Achim, P., Blanchard, X., Piwowarczyk, J.C., Pointurier, F., 2008. Contribution to the development of atmospheric radioxenon monitoring. J. Radioanal. Nucl. Chem. 276, 391–398. Le Petit, G., Cagniant, A., Gross, P., Douysset, G., Topin, S., Fontaine, J.P., Taffary, T., Moulin, C., 2015. SpalaxTM new generation: a sensitive and selective noble gas system for nuclear explosion monitoring. Appl. Radiat. Isot. 103, 102–114. Luke, P.N., 1994. Single-polarity charge sensing in ionization detectors using coplanar electrodes. Appl. Phys. Lett. 65, 2884–2886. Orr, B., Schöppner, M., Tinker, R., Plastino, W., 2013. Detection of radioxenon in Darwin, Australia following the Fukushima Dai-ichi nuclear power plant accident. J. Environ. Radioact. 126, 40–44. Popov, Y.S., Kazarinov, N.M., Popov, V.Y., Rykov, Y.M., Skirda, N.V., 2005. Measuring low activities of fission-product xenon isotopes using the β-γ coincidence method. Instrum. Exp. Tech. 48, 380–386. Prelovskii, V.V., Kazarinov, N.M., Donets, A.Y., Popov, V.Y., Popov, I.Y., Skirda, N.V., 2007. The ARIX-03F mobile semiautomatic facility for measuring low concentrations of radioactive xenon isotopes in air and subsoil gas. Instrum. Exp. Tech. 50, 393–397. Ranjbar, L., Farsoni, A.T., Becker, E.M., 2016. A CZT-based radioxenon detection system in support of the Comprehensive Nuclear-Test-Ban Treaty. J. Radioanal. Nucl. Chem. 310, 969–978. Ranjbar, L., Farsoni, A.T., Becker, E.M., 2017. 135 Xe measurements with a two-element CZT-based radioxenon detector for nuclear explosion monitoring. J. Environ. Radioact. 169–170, 221–228. Ringbom, A., Larson, T., Axelsson, A., Elmgren, K., Johansson, C., 2003. SAUNA—a system for automatic sampling, processing, and analysis of radioactive xenon. Nucl. Instrum. Methods Phys. Res. Sect. Accel. Spectrometers Detect. Assoc. Equip. 508, 542–553. Ringbom, A., Axelsson, A., Aldener, M., Auer, M., Bowyer, T.W., Fritioff, T., Hoffman, I., Khrustalev, K., Nikkinen, M., Popov, V., et al., 2014. Radioxenon detections in the CTBT international monitoring system likely related to the announced nuclear test in North Korea on February 12, 2013. J. Environ. Radioact. 128, 47–63. Saey, P.R.J., De Geer, L.-E., 2005. Notes on radioxenon measurements for CTBT verification purposes. Appl. Radiat. Isot. 63, 765–773. Saey, P.R.J., Bean, M., Becker, A., Coyne, J., d'Amours, R., De Geer, L.-E., Hogue, R., Stocki, T.J., Ungar, R.K., Wotawa, G., 2007. A long distance measurement of radioxenon in Yellowknife, Canada, in late October 2006. Geophys. Res. Lett. 34. SensL, 2019. http://sensl.com/downloads/ds/DS-MicroJseries.pdf. Shilian, W., Qi, L., Qinghua, M., Zhanying, C., Yungang, Z., Huijuan, L., Huaimao, J., Yinzhong, C., Shujiang, L., Xinjun, Z., et al., 2013. Radioxenon monitoring in beijing following the Fukushima daiichi NPP accident. Appl. Radiat. Isot. 81, 344–347. Sivels, C.B., McIntyre, J.I., Bowyer, T.W., Kalinowski, M.B., Pozzi, S.A., 2017. A review of the developments of radioxenon detectors for nuclear explosion monitoring. J. Radioanal. Nucl. Chem. 314, 829–841. Sivels, C.B., McIntyre, J.I., Prinke, A.M., Day, A.R., Lynch, C., Clarke, S.D., Pozzi, S.A., 2019. Stilbene cell development to improve radioxenon detection. Nucl. Instrum. Methods Phys. Res. Sect. A Accel. Spectrom. Detect. Assoc. Equip. 923, 72–78. Warburton, W.K., Hennig, W., Bertrand, J.A., George, S.M., Biegalski, S., 2013. Atomic layer deposition α-Al2O3 diffusion barriers to eliminate the memory effect in betagamma radioxenon detectors. J. Radioanal. Nucl. Chem. 296, 541–549. Zähringer, M., Becker, A., Nikkinen, M., Saey, P., Wotawa, G., 2009. CTBT radioxenon monitoring for verification: today's challenges. J. Radioanal. Nucl. Chem. 282, 737–742.

In this work, a prototype Stilbene-CZT radioxenon detection system is introduced. Pulses are collected in real-time via a digital pulse processor with an onboard FPGA that identifies coincidence events from the CZT and stilbene and transfers it to the PC for processing. The entire detection system is held on a custom designed PCB to reduce noise and enable onboard SiPM mounting. Table 3 shows some of the key figures like energy resolution and memory effect values obtained from this work. Upon injecting the TRIGA irradiated samples in the gas cell, all radioxenon signatures from 131mXe, 133/133mXe, and 135Xe were uniquely identified. A memory effect value of 0.045 ± 0.017% is obtained which is a two orders of magnitude improvement compared to the plastic scintillation cell. The MDC of three of the four radioxenon was found to be below 1 mBq/m3 which is comparable to the state of the art. With improvements in the detection efficiency of the photons by employing multiple CZT detectors around the gas cell, it is estimated that all four radioxenon will have an MDC below 1 mBq/m3. Acknowledgements This work was funded in part by the Consortium of Verification Technology (CVT) under Department of Energy National Nuclear Security Administration award number DE-NA0002534. The authors would also like to express their gratitude to Mr. Adam Grosser (Redlen Technologies) for providing us with CZT crystals. References Alemayehu, B., Farsoni, A.T., Ranjbar, L., Becker, E.M., 2014. A well-type phoswich detector for nuclear explosion monitoring. J. Radioanal. Nucl. Chem. 301, 323–332. Becker, A., Wotawa, G., Ringbom, A., Saey, P.R.J., 2010. Backtracking of noble gas measurements taken in the aftermath of the announced october 2006 event in North Korea by means of PTS methods in nuclear source estimation and reconstruction. Pure Appl. Geophys. 167, 581–599. Bläckberg, L., Fay, A., Jõgi, I., Biegalski, S., Boman, M., Elmgren, K., Fritioff, T., Johansson, A., Mårtensson, L., Nielsen, F., et al., 2011. Investigations of surface coatings to reduce memory effect in plastic scintillator detectors used for radioxenon detection. Nucl. Instrum. Methods Phys. Res. Sect. Accel. Spectrometers Detect. Assoc. Equip. 656, 84–91. Bläckberg, L., Fritioff, T., Mårtensson, L., Nielsen, F., Ringbom, A., Sjöstrand, H., Klintenberg, M., 2013. Memory effect, resolution, and efficiency measurements of an Al2O3 coated plastic scintillator used for radioxenon detection. Nucl. Instrum. Methods Phys. Res. Sect. Accel. Spectrometers Detect. Assoc. Equip. 714, 128–135. Bowyer, T.W., Schlosser, C., Abel, K.H., Auer, M., Hayes, J.C., Heimbigner, T.R., McIntyre, J.I., Panisko, M.E., Reeder, P.L., Satorius, H., et al., 2002. Detection and analysis of xenon isotopes for the comprehensive nuclear-test-ban treaty international monitoring system. J. Environ. Radioact. 59, 139–151. Bowyer, T.W., Biegalski, S.R., Cooper, M., Eslinger, P.W., Haas, D., Hayes, J.C., Miley, H.S., Strom, D.J., Woods, V., 2011. Elevated radioxenon detected remotely following the Fukushima nuclear accident. J. Environ. Radioact. 102, 681–687. Cagniant, A., Le Petit, G., Gross, P., Douysset, G., Richard-Bressand, H., Fontaine, J.-P., 2014. Improvements of low-level radioxenon detection sensitivity by a state-of-the art coincidence setup. Appl. Radiat. Isot. 87, 48–52. Carrigan, C.R., Sun, Y., 2014. Detection of noble gas radionuclides from an underground nuclear explosion during a CTBT on-site inspection. Pure Appl. Geophys. 171, 717–734. Carrigan, C.R., Heinle, R.A., Hudson, G.B., Nitao, J.J., Zucca, J.J., 1996. Trace gas emissions on geological faults as indicators of underground nuclear testing. Nature 382, 528–531. Cooper, M.W., McIntyre, J.I., Bowyer, T.W., Carman, A.J., Hayes, J.C., Heimbigner, T.R., Hubbard, C.W., Lidey, L., Litke, K.E., Morris, S.J., et al., 2007. Redesigned β–γ radioxenon detector. Nucl. Instrum. Methods Phys. Res. Sect. Accel. Spectrometers Detect. Assoc. Equip. 579, 426–430. CTBTO, 2015. https://www.ctbto.org/fileadmin/user_upload/SnT2015/SnT2015_ Posters/T3.1-P7.pdf. CTBTO, 2019. https://www.ctbto.org/specials/who-we-are/. CTBTO Map, 2019. https://www.ctbto.org/map/. CVT, 2017. http://cvt.engin.umich.edu/wp-content/uploads/sites/173/2017/11/11291045-Sivels-Overview-of-Current-Radioxenon-Research.pdf. CVT, 2018. http://cvt.engin.umich.edu/wp-content/uploads/sites/173/2018/10/11_01_ 2018-0940-Sivels.pdf.