Determination of uranium and thorium using gamma spectrometry: a pilot study

This paper presents the results of a pilot experiment aimed at standardizing procedures for the CPqCTR/UESC Gamma Spectrometry Laboratory (LEG) for the quantification of natural radioactive elements in solid environmental samples. The concentrations of 238U, 232Th and 40K in two sediment matrix belonging to the Caetité region were determined, by using the absolute method with uncertainties about 5%. The results were obtained using gamma spectrometry with a high-resolution p-type HPGe detector. As a closure, the absorbed dose, radium equivalent activity and the annual effective dose were calculated.


Introduction
Environmental assessments are one of the permanent activities, in which universities and their research centers are involved, and they can impact on a positive way in a given region. Gamma-ray spectrometry is one of the most widely used techniques for environmental radioactivity assessment [1,2]. It allows to perform a precise quantitative analysis of the radionuclide concentration (natural or anthropogenic) in different types of environmental samples, without the application of laborious sample preparation [3].
However, the quantification of radionuclide in extended sources, by using the absolute method, requires the knowledge of the efficiency detection. The efficiency curve can be experimentally obtained by using a "poly-energetic standard source" with geometry and a matrix composition similar to the studied samples. When these "standard sources" are not available, the efficiency curve for volumetric sources can be obtained from point-like sources of calibration using the efficiency transfer method [4]. Two sediment samples belonging to Caetité region, where it is concentrated nowadays the uranium ore activity in Brazil were analyzed. The efficiency curve determined in [5] was used for quantifying the amount of uranium, thorium and potassium in both samples. This paper presents the results of a pilot experiment aimed at standardizing procedures from the Research Center in Sciences and Technologies of Radiations (CPqCTR/UESC) LEG for the quantification of natural radioactive elements ( 238 U, 232 Th and 40 K) in solid environmental samples.
Based on the fact, that mining is catalogued by the United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR) as a potential source of exposure to natural radioactive elements [6] different radiological magnitudes were also calculated.

Sample collection and preparation
The sediment samples (SS1 and SS2) were collected from two specific sites of the influence region of uranium mining in Caetité, Bahia. The samples were stocked and sealed in plastic bags in order to avoid any external contamination. After that, the samples were sent to our Environmental Samples Preparation Laboratory at the CPqCTR and prepared according to the procedure established in [7]. The granulometric reduction of the collected samples was done using a sieve of 170 mesh (90 µm). After the homogenization procedure, samples were dried in the oven at a temperature of 110 ºC. At intervals of 2 hours, the mass of the samples was verified, until achieving a stable difference (lesser than 5 mg) in the dry weight, in temperature and humidity conditions about 20 ºC and 40%, respectively. The dried samples were hermetically sealed in two plastic containers (vessels) with a cylindrical geometry (6.2 cm 2 x 2.7 cm), previously identified. Finally, both samples were stored during 30 days in a desiccator in order to reach the secular equilibrium of radon with other elements of the natural radioactive series.

Experimental Setup
The experimental measurements were carried out using a coaxial p-type HPGe detector (model GC2518), manufactured by Canberra. The geometrical dimensions and spectrometric characteristics of the detector were verified by [5] and compared with those provided by the manufacturer. This detector has a nominal relative efficiency about 25% compared to 7.62 cm x 7.62 cm NaI (Tl) detector. The measured resolution (FWHM) and the Peak-to-Compton ratio (P/C) for the 60 Co gammaray energy of 1.33 MeV were 1.72 keV and 58:1, respectively. These characteristics were in disagreement with those reported by the manufacturer for the same energy (1.62 keV and 61.6:1, respectively). The detector is installed in a 10-cm-thick lead shield, model 747 [8]. The HPGe is connected to a pre-amplifier (model 2002C) and a Digital Spectrum Analyzer (DSA-200), with a shaping time of 4 µs. The course and fine gain were fixed in 40x and 1.0502x, respectively. A high voltage of 4000 V was applied to the detector.
The energy calibration function, which establishes the relation between gamma-ray energies and the electronic channels, was obtained in the energy range from 59-1408 keV. Adjustment of the experimental data was achieved by an excellent linear fit (r 2 =0.999), whose parameters are: m=0.2302 keV channels -1 and b= -0.6077 keV.
The full-energy peak efficiency calibration was performed in the same energy range with point-like standard sources of: 22 Na, 60 Co, 133 Ba, 137 Cs, 152 Eu, 155 Eu and 241 Am. The spectrum acquisition time was chosen depending on the activity of the source to keep the statistical uncertainty in the peak area below 1%. Background radiation was measured for 72 h, and its contribution was subtracted from the corresponding standard source spectra.

Results
The samples were measured at 5 cm from the top of the detector for 72 h. Spectra were acquired using the software Genie 2000 [9]. Gamma-ray energies of 351 keV ( 214 Pb), 1120 and 1764 keV ( 214 Bi) were used for 238 U calculations. The energies of 583 keV ( 208 Tl) and 968 keV ( 228 Ac) were used for 232 Th calculations, and for 40 K quantification was used its own gamma-ray energy (1460 keV). The measured spectrum of the sample SS2, is shown in figure 1. Due to the analytic lines, used in the quantification procedure (highlighted in red), were above 350 keV, the spectrum is shown for energies greater than 300 keV.
The specific activity of 238 U, 232 Th and 40 K, corresponding to both sediment samples from Caetité region, were obtained using the Absolute Method (Eq. 1): where CPSs and CPSg represent the count rates in the full-energy peak and the background count rates, is the efficiency corresponding to energy E of interest, represents the photon yield or photonemission probabilities to energy E, and m is the dried mass of the sample in kg.
Energy and photon yield values were taken from the database of the International Atomic Energy Agency [10].Background spectra were used for the calculation of the detection limits (DL) and the Minimum Detectable Activity (MDA). In this work was used the Currie modified criterion for the calculation of detection limit and therefore to MDA [11]. The efficiency curve was obtained after adjusting the nominal detector parameters using a Monte Carlo N-Particle eXtended (MCNPX) transport code [12], and using the efficiency transfer method. Correction factors such as solid angle for the volumetric sample and self-absorption correction factors (Fself-abs) were also taken into account to calculations. If the reader is interested, additional details can be found in [5]. Figure 2 shows the efficiency curve in the energetic range of 53-1764 keV for the studied volumetric sample. Red points represent the efficiency values for the analytical gamma lines of interest.  The 238 U, 232 Th and 40 K concentration values (table 1) were estimated from the specific activity Eq. (1), with a mean relative error of 5%. Due to the spectrum acquisition time was chosen to keep the statistical uncertainty in the peak area below 1% (Type A uncertainty), and the photon yield of this lines are lower than 0.3 % (Type B uncertainty); the concentration values' uncertainties were given mainly by the uncertainties in the estimative of the efficiency. For all the efficiency values, the uncertainty was computed as the quadratic sum of the individual uncertainties of the factors used in the efficiency calculation. These factors and their relative errors, are listed in table 2.  As a closure for this work, radiological protection magnitudes such as: the absorbed dose rate (D), the radium equivalent activity (Raeq) and the annual effective dose equivalent (H) for 1 m from the surface for adults were calculated according to [13].
The annual effective dose values do not exceed the accepted as safe limit for occupationally exposed persons (20 mSv y -1 ) [13]. They are even included in the radiation dose limits to mine workers at different uranium operations around the word (< 4.8 mSv) [14].

Conclusions
The pilot experiment carried out in the CPqCTR/UESC LEG established a methodology to perform quantitative procedure of radionuclides by using the Absolute Method. It successful application allowed to determine the concentration of 238 U, 232 Th and 40 K in two sediments samples belonging to the Caetité region and the assessment of the outdoor dose, which do not exceed the regulated levels. This work itself do not constitute an environmental study, because there are not enough sampling points for characterizing the region.