Development and characterization of a mobile γ spectrometer and its field deployment for in situ radioactivity measurements

A mobile γ–ray spectrometer (AMESOS) has been developed using a 3”×3” NaI(Tl) detector, a custom–made mounting holder, and portable electronics to perform in situ measurements of radioactivity. The spectrometer was calibrated using standard point sources and its absolute efficiency was determined. As a field test operation, AMESOS was deployed on the premises of the University of Athens Zografou campus focusing on estimating the NORM levels. Data were analyzed and used to create radiological maps for the metropolitan UoA campus for the first time. Besides natural radioactivity levels, trace concentrations of 137Cs were also detected, attributed to the Chernobyl fallout in Greece. An overall steady performance of the spectrometer was observed throughout the field operation. AMESOS is ready to be deployed for in situ studies of environmental radioactivity and radwaste management.


Introduction
Mobile γ-ray spectrometry is a well -established technique in environmental radioactivity measurements, which can provide results in harsh or remote environments (ICRU 1994), especially when sampling is difficult or when information is needed in a very short time scale. Nowadays mobile spectrometers are rather simple to deploy in the field and can be carried by humans (Kock and Samuelsson 2011, Cresswell et al. 2013, 2018, mounted on robotic instruments (Jilek et al. 2016), air vehicles (Simon andGraham 1998, Kock andSamuelsson 2011), drones (Kurvinen et al. 2005, Pöllänen et al. 2009), or dive into deep oceans aboard marine submersibles (Hattori 2000, Hattori et al. 2001, Tsabaris et al. 2011, Androulakaki et al. 2016. Depending on the application, different types of detectors can be used, such as scintillators (e.g. NaI(Tl)) or solid-state detectors (e.g. HPGe), coupled to integrated electronics modules for data acquisition or custom-made designs to accommodate special needs. The major drawback of mobile systems has been the limitations in supplying stable, uninterrupted power, which generally shortens the duration of measurements in the field. However, as technology leaps forward, power autonomy becomes less and less of an issue, thus extending the range of applications where a mobile spectrometer can be exploited. Scintillators are typically favored over solid-state detectors (Ge(Li) or HPGe) for mobile spectrometers, since the latter traditionally require cooling with liquid nitrogen to operate, are generally more expensive, and are overall more sensitive in field deployment, despite the higher resolution they offer (Clouvas et al. 2004). On the other hand, scintillators are sensitive to environmental conditions, especially temperature changes that may cause gain shifts during operation (Gilmore 2008).
In any case, a careful characterization of the spectrometer is required, especially in terms of its absolute efficiency. The response function of the spectrometer depends on the type of the detector, the geometrical characteristics of the spectrometer and the energies of the photons emitted by the isotopes of interest. Standard point sources are typically used for γ-ray spectrometers, but for mobile systems aiming at measurements in wide-open areas, a characterization under realistic conditions is often additionally necessary.
In this paper, a mobile spectrometer, AMESOS: A MobilE Spectrometer for environmental radioactivity Studies 1 , has been developed in the National and Kapodistrian University of Athens (UoA) to support a basic program of in situ environmental radioactivity measurements (see photo of the actual setup in Fig. 1. So far, there has been no such spectrometer in the pool of 1 the name reflects to the Greek word ΑΜΕΣΟΣ: immediate, direct instruments at the Nuclear Physics Laboratory at UoA and AMESOS is expected to significantly expand the in-house research program in environmental radiation studies. The technical aspects of the spectrometer and its full characterization prior to field deployment are described later in the text. As a proof of good operation in the field, AMESOS was used to measure the natural radioactivity levels in the UoA Zografou Campus, the largest university campus in Greece, hosting ≈70,000 students and staff. To the best of our knowledge this study has been carried out systematically for the first time. The results are presented and discussed in the respective section.

Description of the detection apparatus
AMESOS utilizes a 3"×3" NaI(Tl) crystal coupled to an ORTEC digiBASE 14-pin PMT base with integrated bias supply, a preamplifier, and a MCA with digital signal processing (see ref. ORTEC 2020). The digiBASE ensures low power consumption and gain stability, as well as it allows for easy communication with a dedicated portable PC equipped with the MAESTRO32 software via USB.
The software enables online analysis, while it also allows for nuclide identification and photopeak integration during the measurement. Typical settings of operation of AMESOS during the field measurements and the lab tests are +700 V for the HV bias, 1024 channels and spectrum energy range ≈ 0-2000 keV.
The detector is placed vertically (Fig. 1), with its end cap facing the ground, inside a custom-made hollow aluminum holder of cylindrical shape (wall thickness is 3 mm), standing on a 3-mm thick plexiglass window mounted firmly at the bottom end of the holder. An additional 3-mm thick lead jacket fills the void space between the detector and the holder and is used as a thin radiation shield from the side. Three removable, extensible legs are mounted on the outer surface of the holder, forming a tripod, which allows for positioning the detector at various heights up to 160 cm from the ground. Horizontal alignment is checked with a bubble meter before starting a measurement. A handheld GPS device is used to track the exact coordinates of the setup for each measurement. The whole apparatus is lightweight and can be easily carried around in a backpack or a carriage.
Recorded spectra can be analyzed offline with the specialized software codes SPECTRW (Kalfas et al. 2016) and XSA (Kumbartzki 2018). In such a configuration, the laptop's power autonomy is the only potential limiting factor of AMESOS operation, which can be overcome easily if long-life battery packs or solar cells are used. During operation, each measurement in the field was performed for a time interval of ≈3 h, which was proven sufficient to run without any power outage interruptions.

Characterization of the spectrometer
Prior to field deployment, the spectrometer was fully studied in terms of its calibration, linearity, energy resolution and absolute efficiency. Separate sets of measurements were carried out to study these properties, as described below.
Lacking an extended-area radionuclide standard, such as those offered by IAEA (IAEA 1987), the calibration was carried out using point sources available at the UoA Nuclear Physics Laboratory. Three standard point sources of known activities, i.e. 22 Na (511 and 1274.54 keV), 60 Co (1173.23 and 1332.49 keV), and 137 Cs (661.66 keV) (Dulieu et al. 2017), were individually placed directly below the detector at point "O" (see Fig. 2). The detector was place at a height h = 1.60 m from the ground.
Spectra for each source were collected for approximately one hour in this series of measurements. The first step in characterizing the spectrometer was to perform an energy calibration and check on its linearity. For the calibration, besides the five photopeaks from the sources, two additional summing peaks (1785.54 keV / 22 Na and 2505.72 keV / 60 Co, respectively) were used, reaching the typical energy range needed for environmental radioactivity studies (up to ≈3000 keV). The energy values of the photopeak centroids are plotted in Fig. 3a as a function of the respective channel numbers recorded in the raw spectra. A 2 nd -order polynomial function was used to fit the data and test AMESOS linearity. The fitting curve is also shown in Fig. 3a. The parameters a 0 , a 1 and a 2 of the fitting function: E = a 0 + a 1 * (chn) + a 2 * (chn) 2 (1) with (chn) being the channel number, are shown in Table 1. The value of a 2 is very small, reflecting on the good linearity of the spectrometer in the selected energy range. From the same spectra, the energy resolution of the detector was determined in terms of the full width at half maximum (FWHM) as a function of photopeak energy, E. The experimental data are shown in Fig. 3b together with a fitted 2 nd -order polynomial function of the form: An overall good fit was produced resulting in the parameter values shown in Table 1. Table 1. Values of the fitted parameters for eq. 1 (top rows), eq. 2 (mid rows), and eq. 4 (bottom rows), respectively  Further, the absolute efficiency, ε(E), of the spectrometer was further deduced, using equation (Gilmore 2008, El Khatib et al. 2016: where Ν meas is the measured counting rate of the detector for a particular γ-ray photopeak of energy E, corrected for dead time (less that 0.1% throughout the measurements); N o is the γ-ray emission rate from the source, and P γ is the relative emission probability of a particular γ-ray with energy E. Uncertainties include the statistical error from the counting rate, the uncertainty of P γ found in evaluated nuclear databases (Dulieu et al. 2017), and a 3% relative error in the initial activity of the calibration sources, as provided by the manufacturer. Experimental data points are plotted as a function of energy to extract the absolute efficiency of the detector (Fig. 4).
The efficiency data were fitted with Eq. (4), a simplified version of Eq. (1) in McNelles and Campbell (1973). The results are shown as a function of energy E (in keV) in Fig. 4.
The resulted values of the fit parameters are also shown in Table 1.
A final step before field deployment is finding the correspondence between the recorded photopeak counting rate (counts per second, cps) to spatial activity [Bq m -2 ]. For this task, a circular area of uniform surface distribution with a radius of approximately 2.4 m can be assumed as the source for all counts registered in the detector, for the particular height used in the present study (h = 1.6 m). This radius value has been found by extrapolating the results by Billings et al. (2003). Using a source area of S = πR 2 = 18.25 m 2 , all activities measured in AMESOS can now be converted to spatial activities.

Field measurements
After characterizing the spectrometer, a series of measurements were performed in the field. As a proof of good operation, AMESOS was deployed on the premises of the metropolitan Zografou campus of UoA. The metropolitan campus covers the largest area among university campuses in Greece and hosts about 65,000 students and 3,000 personnel on a daily basis (UoA 2020). The campus is located at the western foot of Mt. Hymettus in Athens and has a large percentage of undisturbed land covered by grass, bushes and trees. (Fig. 5). The area mainly consists of a) limestones aged Upper Cretaceous (Cenomanian), which are intensely fissured and traversed by a dense and chaotic network of secondary calcite veins infused with  Fe-oxides, b) Pleistocene terrestrial and fluvial-terrestrial deposits of various lithological composition (IGME 1996). To the best of our knowledge, a systematic study of the natural radioactivity levels has never been performed on its premises, providing additional motivation to deploy AMESOS at several locations and create radiological maps of naturally occurring radioactive materials (NORM), and of persisting 137 Cs concentrations -if anydue to the Chernobyl fallout in 1986.
In situ measurements were performed at twenty (20) different locations of the campus, marked on the campus map shown in Fig. 5. Undisturbed forest lands, as well as parking lots, building backyards etc, were selected as measurement locations, so as to obtain a representative radiological map of the area in the end. A typical spectrum is shown in Fig. 6. The duration of measurement was kept at Δt≈3 h per location. The corresponding GPS coordinates are shown in Table 2 together with the corresponding spatial activities for 226 Ra, 40 K and 137 Cs. All data were combined further to construct radiological maps.
The distribution maps for 226 Ra, 40 K and 137 Cs are shown in the Fig. 7 top, middle and bottom panels, respectively. 226 Ra content is estimated from its daughter nucleus 214 Bi, which has a half-life of t 1/2 = 19.8(1) min and is one of the most important γ emitters in the 238 U decay chain. In this study, it has been assumed in secular equilibrium with the longer lived 226 Ra (t 1/2 = 1600(7) a). The 226 Ra spatial activities are in the range of 0.240-0.749 kBqm -2 with a weighted average value of 0.478(2) kBqm -2 .
The spatial activity of 40 K falls between the values of 4.64 kBqm -2 and 13.87 kBqm -2 with a weighted average value of 7.10(1) kBqm -2 , which is in fair agreement with a reported value of 9.5(3) kBqm -2 (Vosniakos et al. 1997) and others (Anagnostakis et al. 1996, Clouvas et al. 2004. Early studies of potassium in Earth's crust report spatial concentrations of 40 K ranging 0.51-15 kBqm -2 (Eisenbud 1973), in good agreement with the present results.
On the other hand, measurable 137 Cs activities have been recorded in only half of the locations studied, all being locations with undisturbed soils. It is known that soon after fallouts, radiocesium is trapped in the top layer of soils and sediments at depths not exceeding 5 cm (Fawaris and Johanson 1995, Isaksson and Erlandsson 1995, Ioannides et al. 1996, Giannakopoulou et al. 2012, Kato et al. 2012). Table 2. Spatial concentrations in kBq m -2 for NORM 226 Ra, 40 K and the artificially produced 137 Cs at the various locations measured in the present study. 226 Ra was extracted from its daughter 214 Bi. Half of the locations showed no 137 Cs distribution (marked as bdl -below detection limit). The experimental uncertainties (shown in parentheses) reflect on the statistical error of the counting rate. All locations are forest/park/low-vegetation areas except those marked with (*) which correspond to parking lots.  Besides the natural decay of radiocesium, any source of soil disturbance may result in additional reduction of 137 Cs activity. An additional motivation behind the initial plan to deploy AMESOS in UoA was to target such undisturbed soils, in an attempt to study the persistence of Chernobyl fallout in soils of Athens and elsewhere. Athens with its neighboring suburbs form a megacity of 5 million people, where undisturbed land is nowadays rare to find and study in terms of the long-term impact of man-made radioactivity. In that aspect, the metropolitan UoA campus is a great natural lab to assess the situation, almost 35 years after the 1986 Chernobyl fallout in Greece. The 137 Cs values are significantly lower than those recorded in Athens immediately after Chernobyl (Simopoulos 1989, Kritidis et al. 1990 or later (Kritidis and Florou 1995, Anagnostakis et al. 1996, Petropoulos et al. 1996, showing a significant reduction of its retention in soils. The present 137 Cs measurements fall in the range 0.211-0.511 kBqm -2 and are ≈20 times lower than the average value of 9.0 kBqm -2 in Greek soils reported in 1990 by (Kritidis et al. 1990) and ≈10 times lower than the average 4.0(2) kBqm -2 reported for Attica prefecture 10 years after Chernobyl fallout (Vosniakos et al. 1997). The reduction may be attributed to environmental processes, other than the natural decay of radiocesium, such as soil erosion, surface runoff, migration to deeper soil layers and plant root uptake Florou 1995, Arapis andKarandinos 2004). This reduction is also in agreement with recent studies of undisturbed soils in locations of Northern Greece (Mertzimekis et al. 2014) where the Chernobyl fallout peaked in Greece (Simopoulos 1989, Kritidis et al. 1990, ICRU 1994, Kritidis and Florou 1995, Vosniakos et al. 1997, Arapis and Karandinos 2004, Clouvas et al. 2004).

Conclusions
A mobile spectrometer, AMESOS, was developed and characterized at the University of Athens to be used for environmental radioactivity studies. The spectrometer was found to behave well during field operation performing γray spectroscopic studies. The first application of the spectrometer focused on assessing the NORM distributions ( 226 Ra and 40 K) in the National and Kapodistrian University of Athens Zografou campus for the first time. In addition to the NORM levels, the man-made 137 Cs levels were examined in the studied locations and were found above detection limits in half of them. Despite the respective areal concentrations are small, the effects of the Chernobyl fallout for more than 30 years in the undisturbed soils of Athens seem to persist. Based on the measurements, a set of radiological maps were constructed for the UoA campus to serve as reference for future studies in the area. Overall, the deployment of AMESOS in the field can be considered successful. The mobile spectrometer will be further used for both research and educational purposes in the near future.

Data availability statement
Raw gamma-ray spectra collected during this research are openly available at: https://doi.org/10.6084/m9.figshare.14794449. All other derived data are available from the corresponding author on reasonable request.