A practical experimental approach for the ...

Nuclear Instruments and Methods in Physics Research A 763 (2014) 132–136

Contents lists available at ScienceDirect

Nuclear Instruments and Methods in Physics Research A journal homepage: www.elsevier.com/locate/nima

A practical experimental approach for the determination of gamma-emitting radionuclides in environmental samples Andra-Rada Iurian a,b,n,1, Constantin Cosma b a b

Faculty of Environmental Science and Engineering, Babeş—Bolyai University, Cluj-Napoca, Romania Interdisciplinary Research Institute on Bio-Nano-Science of Babeş—Bolyai University, Cluj-Napoca, Romania

art ic l e i nf o

a b s t r a c t

Article history: Received 1 March 2014 Received in revised form 23 May 2014 Accepted 9 June 2014 Available online 18 June 2014

The work presents a practical and cost saving approach for the experimental efficiency calibration of a coaxial n-type high-purity germanium detector (HPGe), with the intended use in routine laboratory measurements of all gamma emitting radionuclides in different environmental volumetric samples. The calibration was performed using the 238U and 232Th natural radionuclide chains in secular equilibrium, and additionally KCl salt of analytical purity, covering the energy range between 26 and 1461 keV. The performance evaluation in the worldwide proficiency test (PT) organized by the International Atomic Energy Agency in 2013 confirmed the reliability of the analytical measurement results provided by this experimental approach for both natural and artificial radionuclides. Yet, further consideration is needed in the low energy range regarding the effect of the sample composition on the self-absorption and geometry corrections. & 2014 Elsevier B.V. All rights reserved.

Keywords: Efficiency calibration Gamma detector Radionuclide Self-absorption correction Proficiency test

1. Introduction Gamma-ray spectrometry is a widely used, non-destructive measurement technique for the analysis of different gammaemitting radioisotopes in environmental samples. Standard point sources having single energy emissions or volumetric sources of mixed radionuclides, obtained through primary methods and commercially available are commonly employed for efficiency calibration of the gamma detectors. An example of multinuclide standard solution would consist of artificial radionuclides such as 241 Am, 109Cd, 57Co, 139Ce, 203Hg, 113Sn, 85Sr, 137Cs, 60Co and 88Y, with energy lines in a wide spectra region (59–1836 keV). The great advantage of using this mixed-radionuclide standard source is that no correction must be performed for interfering radionuclides and nor for the background. Hence, a disadvantage in the short halflives of several nuclides must be recognized, thus making the source useful over the whole spectra region only for a short period of a few months, being also very costly for the radioanalytical

n Corresponding author at: Faculty of Environmental Science and Engineering, Babeş—Bolyai University, Cluj-Napoca, Romania. Tel.: þ 40 264 307 032; fax: þ40 264 307 030. E-mail addresses: [email protected], [email protected] (A.-R. Iurian). 1 Current address: Terrestrial Environment Laboratory, IAEA Environment Laboratories, Department of Nuclear Sciences and Applications, International Atomic Energy Agency, Vienna International Centre, PO Box 100, 1400 Vienna, Austria.

http://dx.doi.org/10.1016/j.nima.2014.06.032 0168-9002/& 2014 Elsevier B.V. All rights reserved.

laboratory. The half-life of any mixture like this is given by the halflife of its shortest-lived member. Another example of multigamma standard would include 125 Sb, 154Eu and 155Eu, covering the energy range between 27 and 1596 keV and having half-lives beyond 2.5 years [1]. The disadvantage of this mixture is the many cascade transitions of Eu for which high coincidence summing corrections are needed, giving additional uncertainties in the massic activity results. The use of a radioactive chain source in secular equilibrium can be seen as a practical approach for the efficiency calibration of a gamma system, as its half-life is given by the longest-lived parent. U and Th are the only very long-lived radionuclides with natural radioactive chains. Yet, they have a disadvantage in the lower specific activities, leading to the necessity of physically larger volume sources. This study presents the experimental efficiency calibration of the HPGe detector from Cluj-Napoca laboratory, using RGU-1 and RGTh-1 certified materials provided by the International Atomic Energy Agency (IAEA). These materials were prepared in 1987 in a different manner than the previous ones, being considered to form a set of three certified sources (including RGK-1) to be used for the calibration of laboratory gamma-ray spectrometers for U, Th and K measurements in geological samples [2]. They were prepared using natural U and Th ore, respectively, diluted with floated silica powder, and contain 238U and 232Th natural radionuclide chains in secular equilibrium. Instead of RGK-1, our calibration method included KCl salt crystals of analytical purity, provided by Merks.

A.-R. Iurian, C. Cosma / Nuclear Instruments and Methods in Physics Research A 763 (2014) 132–136

Disequilibrium state in the decay chain 226Ra–210Pb, caused by the gaseous 222Rn escaping from the source container after a long time period (420 years) since the preparation and enclosing of the material, was taken into account. The calibration procedure is presented with the aim to be used within the laboratory for routine activity measurements of gamma emitting radionuclides of unknown energies. It has the advantage in allowing the determination of any gamma-emitting radionuclide in different liquid and solid environmental samples, even if not present in the calibration source. The proposed efficiency calibration procedure and the laboratory performance in radioanalytical measurements was verified through participation in the IAEA-TEL-2012-03 worldwide open proficiency test on the determination of radionuclides in water, hay and soil, organized by the IAEA Laboratories in 2013. The final results and corrective actions are presented further on.

2. Materials and methods 2.1. Gamma-ray spectrometer An Ortec n-type coaxial detector (GMX30P4-ST) with Be window, 34% relative efficiency and a resolution of 1.92 keV at 1.33 MeV line of 60Co has been employed for the gamma measurements. It is connected to a Digital Multichannel Analyser and mounted on mechanical cooling system. The gamma-ray spectrometric facility also incorporate passive background shielding, using a non-contaminated old lead tower of 8 cm thickness with additional Cu shield inside the tower, of 3 mm. Gammaspectra were analysed with dedicated commercial software for gamma-ray spectrometry, namely Maestro-32. The detector resolution and energy calibration is periodically verified for stability using a water-equivalent 152Eu source [3] and a 241Am point source, certified by the Horia Hulubei National Institute of Research & Development for Physics and Nuclear Engineering Bucharest (Romania). 2.2. Calibration sources Three samples containing RGU-1, RGTh-1 and KCl were prepared and measured with the n-type Ortec detector. The calibration sources were encapsulated in aluminium containers (80.5 mm diameter, 0.22 mm bottom thickness, 1.00 mm side wall thickness and 26.5 mm filling height) which do not allow radon diffusion [4]. After properly sealed with glue and electrical band, the calibration samples were kept for minimum four weeks before measuring to ensure that equilibrium between 222Rn and 226Ra is reached. The container geometry was selected with a diameter close to the end cap diameter of the detector. The measurement time was set to 24 h for RGU-1 and RGTh-1 and to 4 h for KCl, in order to reach meaningful statistics for each full energy peak which meant to be used for efficiency calibration. This measurement time gave a dead time correction of 0.5%. The spectra were collected separately and afterwards analysed and plotted into the same file to cover the whole energy range. RGU-1 contains U-ore with 400 ppm of U, less than 20 ppm of K and less than 1 ppm of Th and RGTh-1 contains 800 ppm of Th, 6.3 ppm of U and K (0.02%) [2]. The specific activity concentration (Bq g  1) of 238U, 235U and 232Th and their natural abundance were used to derive the radionuclides activity (Bq). Thus, the specific activity of 238U was found to be 4.939970.0250 Bq g  1, 0.23037 0.0012 Bq g  1for 235U and 3.253670.0650 Bq g  1for 232Th. In the case of 40K, its specific activity was determined considering the elemental weights for KCl and 40K natural abundance. It corresponds to a 40K specific activity of 2.645  105 Bq kg  1. All nuclear

133

data used in the gamma-spectrometric analysis were extracted from Monographie BIPM-5 [5]. 2.3. Reference materials and proficiency test samples The internal validation of the efficiency calibration curve was performed with three IAEA reference materials: IAEA-447 soilmoss [6], IAEA-375 soil [7], IAEA-372 hay [8] enclosed in cylindrical plastic containers of 90 mL (75.98 mm external diameter, 0.82 mm side and bottom wall and 29.00 mm height) and measured for more than 24 h. The selected materials included different radionuclides and matrices, in order to cover a wide energy range. Within the worldwide proficiency test (PT), each participating laboratory received five homogenized samples of different matrices (3 water bottles, one soil box and one hay box) for the analysis of the radionuclide content. The water was spiked with unknown anthropogenic radionuclides and with 226Ra to simulate a natural background. One water sample contained known radioisotopes and massic activities, and was provided for quality control purpose. The hay sample contained hay collected in Austria, shortly after the Chernobyl accident, spiked with 134Cs and ‘naturally contaminated’ with 137Cs. Soil sample contained unknown natural and anthropogenic radionuclides and was collected from Jeju Island, Republic of Korea. All results were to be reported in Bq kg  1, on a dry-weight basis. The samples were prepared in 90 mL cylindrical plastic geometries and measured for 24 h, apart from the soil sample which was measured for 72 h in order to achieve statistically meaningful data. All samples were measured under the same conditions as those under which the system has been calibrated. The final results were given as massic activity7combined standard uncertainty (Bq kg  1, dry mass), 95% confidence interval. The uncertainty budget followed the error propagation low and included the following components: sample and calibration source counting statistics, calibration source activity, background correction, photopeak efficiency fitting, self-attenuation and geometry correction and coincidence summing correction. 210Pb was determined from its low gamma-energy at 46.5 keV, the 1460.8 keV energy line was considered for 40K estimation and 228Ac was determined from its 911.19 keV gamma-line. 208Tl and 212Pb were determined from the 583.19 keV and 238.63 keV gamma energies, respectively. 137Cs was calculated from the 661.6 keV gammaemission of 137mBa and 241Am was estimated from its 59.54 keV line. No correction has been made for ingrowth from 241Pu. Two or three gamma-energies were analysed for 134Cs (604.72 keV, 795.86 keV), 60Co (1173.23 keV and 1332.49 keV) and 152Eu (344.28 keV, 778.9 keV and 1408.01 keV) and the resulted massic activities were reported as arithmetic mean. 2.4. Corrections performed for massic activity determination Decay-corrections were applied to calibration sources and samples to be measured with the reference time given in their Certificates or with the one requested by the PT convener. For background corrections, an empty geometry was measured in the same conditions for about 72 h, and the average background net counts of two intercalated spectra were subtracted from the net peak area of the source. Corrections for geometry and self-attenuation, true coincidence summing corrections for calibration sources (for 227Th and 228Ac) and samples (for 152Eu, 134Cs, 60Co, 208Tl and 228Ac) were also considered. The summing correction factors were derived using the available code EFFTRAN [9], while for the efficiency transfer the ETNA code [10] was employed. The use of these codes makes necessary to have a detailed description of the geometric intrinsic characteristics of the detector, which were given by the supplier

134

A.-R. Iurian, C. Cosma / Nuclear Instruments and Methods in Physics Research A 763 (2014) 132–136

based on request. Also, the geometry characteristics of the sample container and sample chemical composition were needed. The elemental composition of RGU-1 and RGTh-1 is known and it is available in the Reference Sheet given by the IAEA [2]. For soil samples, a standard soil composition (O: 55.84%, C: 1.2%, H: 1.11%, Al: 7.10%, Si: 31.55% and Fe: 3.2%) was used, while the hay was considered as cellulose (C6H12O6).

3. Results and discussions 3.1. Efficiency calibration The efficiency calibration procedure included: (i) true photopeak efficiency determination, (ii) background measurements, (iii) coincidence summing corrections, and (iv) internal validation using IAEA certified reference materials. The apparent photopeak efficiency was derived from the radioactive decay equation, considering the known source activity (Bq), the net counts and the radionuclide yield in the spectra for a certain gamma-energy. In the case of the overlapping peaks of the same radioactive chain in secular equilibrium, the sum of all decay-probabilities was considered. It must be pointed out that the low energy of 210Pb from 46.5 keV was not included in the calibration curve due to the disequilibrium state in the 226Ra–210Pb decay chain. Moreover, the photopeaks of 222Rn and 220Rn progenies and the ones strongly affected by coincidence summing were not included in the efficiency curve. The efficiency fitting curve for the GMX detector is given in Fig. 1 as a 4-order polynomial function in natural logarithm scale using the least-squares method. The true photopeak efficiency of the GMX HPGe detector was measured in a range between 25 keV and 1461 keV, using three different natural radioactive sources and identically geometry containers. The efficiency curve shows two regions of different behaviour because distinct attenuation and absorption processes are involved. The efficiency rises rapidly at low energies due to the high attenuation effect, until around 100 keV and slowly decreases again afterwards. As consequences, two calibration curves were fitted: one in the low energy range (o250 keV) and the second in the high energy scale (4250 keV) of the spectrum, with a common point between. This method allowed us to obtain a better fitting for both energy ranges. The differences between the experimental efficiency and the fitted efficiency for several selected gamma-energies used in the calibration were less than 1.6% for the low-energy range (o250 keV), and less than 2.5% for the higher energies (4250 keV), translating in

Fig. 1. The efficiency calibration curve of the n-type Ortec detector (26–1461 keV).

a small contribution in the final combined standard uncertainty of the radionuclide massic activity. Several other authors [11–14] reported the use of IAEA RGU-1, RGTh-1 and RGK-1 reference materials for gamma detector calibration purpose and further on applied the calibration efficiencies to derive the radionuclide activity concentration of soil or sediment samples. Martínez-Ruiz et al. [15] used a dusty Reference Material (phosphate rock) containing high level of 238U-series radionuclides as calibration sample for air filters. Measurement validation represents an important requirement of any laboratory dealing with radioanalytical techniques due to the increasing demand of data quality assessment. The internal validation of the efficiency calibration curve was performed with three IAEA reference materials: IAEA-447 soil-moss, IAEA-375 soil and IAEA-372 hay. The reported differences between the certified and measured massic activities were found within the combined standard uncertainty of the radionuclide activity concentration. In the case of IAEA-447, a difference of 1.5% was determined for 210Pb, 212 Pb and 228Ac relative to their certified activity values, while for 137 Cs and 208Tl around 5% difference was observed. The massic activities of 137Cs and 40K for the IAEA-375 reference material were in 2% disagreement with the certified data, while for the IAEA-372, with different composition (hay) and geometry, a 1% disparity was observed between the measured activities and the certified ones. Thus the experimental calibration curve was internally validated for the entire spectra region and also for different sample matrices. It has been shown that certified materials with radioactive chains in secular equilibrium can be used for the calibration of the gamma spectrometric systems if all necessary corrections are being performed. The main advantage is that the whole energy region is covered by the calibration curve, extending the number of radionuclides that can be measured, even if not present in the calibration source. Still, the stability of the new efficiency calibration based on full-energy peaks of natural radionuclides must be ensured by observing and controlling the variation of detector background. The integral count rate in the background spectra over the period of the measurements had a value of 3.49 cps. Dragounová and Rulík [16] showed through long time series of background measurements that low activity determination of natural radionuclides can be influenced by background fluctuation. However, these authors also concluded that the peaks of 210Pb, 226 Ra (at 186.2 keV) and 40K, as well as integral count rates do not show significant background fluctuations.

3.2. External validation through participation in the IAEA-TEL-201203 worldwide open proficiency test Both internal and external validation procedures are an important part of good laboratory practice and also a requirement of the accreditation body for in-house methods [17,18]. This represented the first participation of our laboratory in a worldwide proficiency test employing gamma measurements of environmental samples with different matrix (water, hay, soil). The performance of the participant laboratories was evaluated for accuracy and precision relative to the target values established by the IAEA. Table 1 shows the analytical performance evaluation of the IAEA PT in various matrices. Most of results were slightly underestimated compared with the target values given by the IAEA in the Laboratory Report, but still within the Maximum Acceptable Relative Bias (MARB). The laboratory had a normalized average analytical performance of 92.9% and also showed reasonably good precision evaluated as the relative standard uncertainty. This external evaluation emphasises the validation of the efficiency calibration of the GMX n-type spectrometric system used for the gamma measurements.

A.-R. Iurian, C. Cosma / Nuclear Instruments and Methods in Physics Research A 763 (2014) 132–136

135

Table 1 IAEA-TEL-2012-03 worldwide open proficiency test evaluation. Sample

Analyte

Laboratory value 7 Combined standard uncertainty (Bq kg  1, dry mass)

IAEA value 7Combined standard uncertainty (Bq kg  1, dry mass)

01—Water

134

Cs Eu

79.3 7 6.1 112.8 7 2.6

82.6 76.08 118.6 71.0

Cs Am

103.6 7 7.2 142.6 7 11.2

152

02—Water

137

241

04—Hay

134

Cs Cs

137

05—Soil

40

K Cs 208 Tl 210 Pb 212 Pb 228 Ac 241 Am 137

Relative bias (%)

Maximum Acceptable Relative Bias (MARB)

Final score

 4.04  4.91

15 15

A A

102.5 70.75 120.9 70.74

1.05 17.94

15 20

A A

293.6 7 22.3 786 7 55

316 720 816 724

 7.09  3.68

20 15

A A

183.2 7 7.86 105.3 7 7.31 9.9 7 0.8 7307 64 30.4 7 1.6 32.3 7 1.9 1.357 0.20

207.7 78.3 118.6 72.9 11.5 70.6 595 719 31.0 71.2 32.4 71.6 1.78 70.1

 11.80  11.21  14.00 22.77  2.10  0.28  24.16

20 20 20 20 20 20 25

A A A N A A A

A—‘Accepted’ when both accuracy and precision achieved ‘Accepted’ status. N—‘Not Accepted’ when the accuracy is ‘Not accepted’.

Only in the case of 210Pb, the relative bias was higher than the MARB, even if corrections for geometry and self-absorption were applied and the internal validation check had satisfactory results. Higher relative difference but with accepted status was also reported for 241Am, measured at its low gamma-energy. The soil chemical composition used in the computation of geometry and self-attenuation correction factors was identified as one of the main plausible contributors to the high relative bias for radionuclides measured at low gamma-energies. This information was not given by the PT convener and was not determined experimentally, whereas the composition of the RGU-1 and RGTh-1 was available in their Certificates. The sample geometry and the linear attenuation coefficient (which depends on material density, sample composition, and photon energy) are factors that contribute to the self-attenuation effect in gamma-ray spectrometric measurements [19]. After the PT concluded, the soil matrix composition was determined by X-ray fluorescence spectrometry (XRF) using WDXRF and EDXRF spectrometers from the Faculty of Sciences and Arts, Valahia University of Targoviste, Romania. Carbon has been distinctively analysed using the combustion method. The determined chemical composition of the PT soil is presented in Table 2, by comparison with the standard soil composition initially used in the efficiency transfer computation. The total mass attenuation with coherent scattering was obtained using the XCOM: Photon Cross Section Database [20] application, for both soil compositions given in Table 2. It can be seen (Fig. 2) that high differences between them only appeared in the low energy range; up to 32% at 40 keV. Over 200 keV, the relative bias does not exceed 1.8%. Furthermore, the computation of the efficiency transfer factors (ETF) was performed for the soil sample in the same measurement conditions, using the XRF data. The relative bias (%) resulted between the ETF obtained using the standard soil composition and the one determined by X-ray fluorescence is plotted in Fig. 3. For the 46 keV gamma-energy of 210Pb, the difference between the two ETF values was 15%, while over 250 keV the correction factors were in 99.9% agreement. Our findings showed that complex chemical composition of sample matrices (i.e. soil), involving elements with high atomic numbers, can induce outstanding biases in the radionuclide activity determination at low gammaenergies. Within a Monte Carlo-based hypothetical proficiency test for low energies self-absorption corrections in soil samples with the same density and different chemical composition, Carrazana

Table 2 The chemical composition of the PT soil sample. Compound

O H C Si Al Fe Ca Ti K Mg Na P Mn

Fraction by weight (%) XRF data

Standard soil composition

35.41 – 22.10 15.95 9.88 9.79 2.71 1.28 0.97 0.834 0.431 0.343 0.345

55.84 1.11 1.20 31.55 7.10 3.20 – – – – – – –

Gonzalez et al. [21] found a bias in the activity values between 4% and 70%, with more than 50% of these results being unacceptable. The ETF bias in the low energy range runs into the value of the radionuclide massic activity. Still, it could not be demonstrated that an improvement has been achieved for the 210Pb activity concentration in PT soil sample using the XRF data. Thus, further measurements are needed including the validation of the ETF values in the low energy range by an experimental approach based on the transmission method of Cutshall et al. [22]. Moreover, to exclude any error given by the efficiency curve fitting, the derived true photopeak efficiency values for 210Pb and 241Am must be verified using volumetric sources of single gamma emitters.

4. Conclusions A practical and cost saving approach was described for the efficiency calibration of a n-type HPGe detector. The calibration involved certified natural sources which contained U and Th radioactive series in secular equilibrium, namely the IAEA RGU-1 and RGTh-1, and additionally KCl salt crystals of analytical purity. The internal validation results obtained through gamma measurements of IAEA reference materials (IAEA-447, IAEA-372 and IAEA-

136

A.-R. Iurian, C. Cosma / Nuclear Instruments and Methods in Physics Research A 763 (2014) 132–136

Acknowledgments The work represents part of the PhD thesis of the first author and was performed with support of the POSDRU CUANTUMDOC “Doctoral Studies for European Performances in Research and Innovation” Project-ID79407, funded by the European Social Fund and Romanian Government. Partial support has been also received through the RAMARO project (73/2012). The authors would like to thank Prof. Claudia Stihi from the Faculty of Sciences and Arts, Valahia University of Targoviste, Romania, for proving the XRF measurement of the PT soil sample.

References

Fig. 2. Mass attenuation coefficients for a standard soil composition and for the PT soil composition determined by XRF.

Fig. 3. The relative bias (%) between the efficiency transfer factors (ETF) for the PT soil sample using the XRF data and the standard soil composition.

375) showed very good agreement with the certified values for different matrix materials, using standard chemical composition for the computation of the efficiency transfer correction factors. The external validation through participation in the IAEA-TEL-2012-03 worldwide open proficiency test on the determination of radionuclides in water, hay and soil showed a normalized average analytical performance of 92.9%. Still, corrective actions are further necessary in the low energy range, where the chemical composition of the soil sample highly influenced the computation of the efficiency transfer correction factors. Our work highlights the importance of participating in proficiency tests for the external validation of the analytical laboratory results.

[1] B.M. Coursey, D.D. Hoppes, F.J. Schima, Nucl. Instrum. Methods Phys. Res., Sect. A 193 (1982) 1. [2] A.N. Hanna, J.L. Labrecque, R. Schelenz IAEA/RL/148 1987. [3] M. Sahagia, E.L. Grigorescu, Nucl. Instrum. Methods Phys. Res., Sect. A 312 (1992) 236. [4] D. Pittauerová, B. Hettwig, H.W. Fischer, Radioprotection 46 (6) (2011) 277. [5] BIPM, 2011. Table of Radionuclides, Monographie BIPM-5. Bureau International des Poids et Mesures, Sèvres. 〈http://www.nucleide.org/DDEP_WG/DDEPdata. htm〉. (last accessed on January 2014). [6] IAEA. Reference sheet—Certified Reference Material IAEA-447 Natural and Artificial Radionuclides in Moss-Soil. Vienna, Austria (2011). 〈http://nucleus. iaea.org/rpst/ReferenceProducts/ReferenceMaterials/Radionuclides/IAEA-447. htm〉, (last accessed on January 2014). [7] IAEA. Reference sheet—Reference Material IAEA-375 Radionuclides and Trace Elements in Soil. Vienna, Austria (2000) 〈http://nucleus.iaea.org/rpst/Referen ceProducts/ReferenceMaterials/Radionuclides/IAEA-375.htm〉, (last accessed on January 2014). [8] IAEA. Reference sheet—Certified Reference Material IAEA-372 Radionuclides in Grass. Vienna, Austria (2010) 〈http://nucleus.iaea.org/rpst/ReferenceProducts/ ReferenceMaterials/Radionuclides/IAEA-372.htm〉, (last accessed on January 2014). [9] T. Vidmar, G. Kanisch, G. Vidmar, Appl. Radiat. Isot. 69 (2011) 908. [10] F. Piton, M.-C. Lépy, M.-M. Bé, J. Plagnard, Appl. Radiat. Isot. 52 (3) (2000) 791. [11] M. Psichoudaki, H. Papaefthymiou, J. Environ. Radioact. 99 (2008) 1011. [12] A.D. Simms, C. Woodroffe, B.G. Jones, H. Heijnis, R.A. Mann, J.J. Harrison, J. Environ. Radioact. 99 (2008) 1111. [13] S. Turhan, A. Köse, A. Varinlioğlu, N.K. Şahin, I. Arikan, F. Oğuz, B. Yücel, T. Özdemir, Geoderma 187–188 (2012) 117. [14] J.S. Lee, K.H. Kim, J.H. Shim, J.H. Han, Y.H. Choi, B.-J. Khang, Mar. Pollut. Bull. 64 (2012) 153. [15] F. Martínez-Ruiz, E. Borrego, E.G. San Miguel, J.P. Bolívar, Nucl. Instrum. Methods Phys. Res., Sect. A 580 (2007) 663. [16] L. Dragounová, P. Rulík, Appl. Radiat. Isot. 81 (2013) 123. [17] ISO. Proficiency Testing by Interlaboratory Comparisons—Part 1: Development and Operation of Proficiency Testing Schemes, Guide 43-1 (1997). [18] H.O.F. Andersson, Accredit. Qual. Assur. 3 (6) (1998) 224. [19] E. Robu, C. Giovani, Rom. Rep. Phys. 61 (2) (2009) 295. [20] M.J. Berger, J.H. Hubbell, S.M. Seltzer, J. Chang, J.S. Coursey, R. Sukumar, D.S. Zucker, XCOM Photon Cross Sections Database, NIST, 〈http://physics.nist. gov/PhysRefData/Xcom/Text/XCOM.html〉. (last accessed on April 2014). [21] J. Carrazana Gonzalez, N. Cornejo Diaz, M. Jurado Vargas, E. Capote Ferrera, Appl. Radiat. Isot. 68 (2010) 360. [22] N.H. Cutshall, I.L. Larsen, C.R. Olsen, Nucl. Instrum. Methods Phys. Res. 206 (1) (1983) 309.