Method for the detection of Tc in seaweed samples coupling the use of Re as a chemical tracer and isotope dilution inductively coupled plasma mass spectrometry

1 A method for the detection of Tc in seaweed samples coupling the use of Re as a chemical tracer and ID-ICP-MS. Jose Luis Mas1)*, Keiko Tagami2), Shigeo Uchida2) 1) Dpto. de Fisica Aplicada, Universidad de Huelva, Campus de El Carmen, Facultad de Ciencias Experimentales. 21071 Huelva (Spain) 2) Environmental and Toxicological Sciences Research Group, National Institute of Radiological Sciences, Anagawa 4-9-1, Inage-ku, Chiba-shi, Chiba, 263-8555, Japan. Abstract. The analysis of artificial radionuclide 99Tc in environmental samples requires of a chemical separation due to its low concentration, and therefore the use of a chemical yield tracer is peremptory. From the practical viewpoint, Re could be used for these purposes, due to the chemical similarities between Re and Tc. Thus, the use of a radioactive tracer for Tc recovery calculation could be avoidable. However, results from the more recent intercomparison exercise shows that the use of Re as a chemical yield tracer appears to underestimate the Tc concentration relative to the result obtained with isotopes of Tc. In this work, we show the methodology used for designing a simple separation method for the measurement of 99Tc in environmental samples. Tc recovery is estimated throughout the Re recovery calculation by the ID-ICPMS technique. For chemical separation, a chromatographic resin has been used. Interfering elements are removed using a resin washing step carefully designed to avoid any element fractionation between Re and Tc, being this step of major importance to assure the equivalence of the chemical recoveries for both elements. Such an agreement was tested using 5 replicates of five seaweed samples. The average recoveries for 95mTc and Re were 93±6% and 95±7%, respectively, those were within the uncertainty intervals for each other. The results here explained demonstrated the possibility of applying Re chemical recoveries to calculate the Tc concentrations with the advantage of not introducing systematic errors.  Corresponding author: Fax: (+34) 959 019777. e-mail: [email protected] 2 Keywords: Technetium; Rhenium; Chemical Tracer; ID-ICP-MS. 1. Introduction. Technetium-99 (2.1105 y of half-life) is a low-energy beta-emitter formed as a fission fragment with a relatively high fission yield ( 6%). Controlled radioactive releases from the large nuclear fuel reprocessing plants of La Hague (France) and Sellafield (UK) have been the major sources of this radionuclide in North Atlantic seawater, even in competition with the global fallout source. It will be predominately present in seawater under oxic conditions in the form of TcO4ion, which is extremely stable and it could be carried out far from these sources, following the general circulation of the ocean currents [1]. Indeed, such stable characteristics allow the researchers to use it as a radioactive tracer for the movement of water marine masses and marine currents. Furthermore, several biota species with a high accumulation capacity for this element have been identified [2], such as brown algae, especially those of Fucus Vesiculosus specie. This fact suggested their use as a bioindicator for technetium concentrations in seawater [1]. As a fast and very sensitive alternative to radiometric detection, ICP-MS has been proposed for the 99Tc determination in environmental samples. Hence, quadrupole ICPMS instruments with pneumatic [3-6] or ultrasonic nebulizers [7], Electrothermal Vaporization units (ETV-ICP-MS, [8]), Laser Ablation units (LA-ICP-MS, [9]) and High Resolution ICP-MS [10] have been used for 99Tc detection in a variety of industrial and environmental samples. However, the chemical separation and 3 purification of Tc from environmental samples is needed. To achieve this, a wide variety of isotopes has been used as yield monitors in the determination of 99Tc in environmental samples. Usually, 99mTc [11-12], 95mTc [13] and 97mTc [14] have been used. These isotopes are generally determined using gamma-ray spectrometry. Beals [15] proposed the Isotope Dilution technique coupled with ICP-MS (ID-ICP-MS) using 97Tc as a yield tracer. The main advantage of the ID technique is in overcoming the problems associated with instrumental drift and matrix effects. Furthermore, the IDICP-MS technique relies purely in isotopic ratios; therefore the problem of an incomplete extraction is negated. It is well known that Re and Tc chemistries are similar: their analogous compounds are virtually identical from a structural standpoint. Therefore, Re has been proposed as a tracer for 99Tc radiochemical recovery calculation from environmental samples, both for radiometric [16] and non-radiometric [17] counting methods. The first of these methods [16] follows a chemical separation based on coupling 1) the use of an anion exchange resin to concentrate Tc (and Re), 2) a precipitation step of Tc and Re as sulfides, from a NaClO4 solution, and 3) a precipitation step of Re and Tc as tetraphenyl arsonium salts. Chemical yield is determined by gravimetric analysis of the rhenium salt. The second method [17] is applied after a chemical separation scheme based on the use of a TEVA·Spec chromatographic resin, although unfortunately analytical details are still to be published. Nowadays, Re is used routinely for this purpose with few problems reported. This approach avoids the use of radioactive isotopes, especially at the place where these 4 radioisotopes are strictly forbidden (e.g., on a ship for seawater sampling). Furthermore, using Re avoids the analytical implications of the use of another isotope of Tc: common nuclear reactions for 97Tc and 95mTc generation produce also 99Tc, which cannot be removed from the yield tracer solution [16]. An alternative reaction that produces 95mTc free of 99Tc has been described elsewhere [13], but this option is only available to those laboratories provided with a cyclotron. However, the results from a recent intercomparison exercise for 99Tc in seaweed samples shown that those laboratories who used Re as a chemical yield tracer, systematically underestimated the Tc concentrations by more than 20% on average [18]. This fact could be due possibly to an overestimation of the recoveries for Tc. The main difference between Re and Tc is in their respective oxidation potentials; the reduction of Re is more difficult than that of Tc [19]. Thus, slightly different coprecipitation profiles for ReO4and TcO4ions with TPAC, and different elution profiles from an anion exchange resin have been shown [16]. More recently, the differences in the capacity factors for Tc and Re using the same chromatographic resin have been shown [20]. Those small chemical differences should be considered in detail if a new method that uses Re as a chemical yield tracer for Tc is to be designed. In this work, a new method that uses Re ID-ICP-MS detection for Tc chemical yield calculation is proposed. In this way, data for Tc solution concentration and chemical yield could be achieved using the same analytical running with an only instrument. This methodology offers the following advantages: 1) Radiological concerns due to 5 radioactive tracers handling are avoided 2) ID-ICP-MS improves the precision of the Re chemical yield calculation 3) The method takes into account explicitly the small differences between the Re and Tc chemical behavior. Hence, the problem of the Tc concentration underestimation could be avoided. 4) The ID-ICP-MS technique allows the determination of both Tc concentration and chemical yield simultaneously; therefore additional techniques (such as gamma spectrometry for Tc isotopes recovery calculation or gravimetric analysis for Re recovery calculation) are unnecessary. This paper focuses on the methodological aspects of the separation scheme proposed, and on its justification. The equivalence for Re and Tc recoveries is tested using some seaweed samples. There are no suitable reference materials with 99Tc certified concentrations. Therefore, several seaweed samples from the North Atlantic Ocean were subjected to this purpose. 2. Experimental 2.1. Samples and chemicals Five samples of brown algae (Fucus sp.) were taken on the coast of Utsira, southwest of Norway, in the year 2000 on a monthly basis. Dates of collection are provided on Table 1. Samples were periodically taken in the framework of an evaluation program of environmental radioactivity in the Nordic waters [21]. Five replicates of every sample were analyzed for comparing the results obtained in this work for Re and Tc chemical yields. 6 After the collection, seaweed samples were dried up, and the sediment particles were removed by shaking. Using a mechanical sample homogenizer, the samples were homogenized. Then, the samples were transferred to cylindrical PEEK beakers, and then they were sealed for transportation. Nitric acid used was ultrapure-analytical grade (Tama AA-100; Tama Chemicals Co. Ltd, Kanagawa, Japan). Deionized water (>18 M) was obtained from a Milli-Q water system (Nihon Millipore K.K; Tokyo, Japan). Prepacked columns of Tc-selective chromatographic resin (TEVA, Eichrom Industries, Inc; Sowa Trading Company, Inc., Tokyo, Japan) were used for separation of Tc and Re. The high selectivity of this resin for ReO4and TcO4has already been reported [20,22]. A radioactive tracer, 95mTc (half-life: 61 d, chemical form: TcO4-), was used for 99Tc and enriched 185Re (chemical form: ReO4-) was used for Re. The carrier-free 95mTc was separated from a Nb foil irradiated with alpha particles [13]. The 185Re isotope enriched solution was produced dissolving a suitable amount of the solid supplied by the Oak Ridge National Laboratory (TN, USA) (96.74 ± 0.01 % isotopic abundance of 185Re) to 6M HNO3. Using both yield tracers, it would be possible to compare their respective chemical yields for each sample replicate using isotope dilution technique, as natural abundances are 37.40% (185Re) and 62.60% (187Re). 2.2. Separation Method The samples were oven dried at 65 ˚C. Then, 0.5-g amount of each sample was transferred to a porcelain crucible and spiked with 25 Bq of 95mTc (approximately 30 fg) 7 and 35 ng of the 185Re-enriched solution, respectively. Then the samples were subsequently taken to an electrical oven. Three different sample pretreatment methods were compared, in order to optimize this step: A) the sample was heated at 450˚C for 30 minutes, then, the residue was dissolved using 5 ml of 4M HNO3 at 120˚C for 3 h; B) heating the sample at 450˚C for 90 min, the residues were then dissolved using leaching with 5 ml of 4M HNO3 at 120 C for 3 h; C) heating the sample at 450˚C for 90 min., the residues were then dissolved using leaching with 5 ml of 8M HNO3 at 120 C for 3 h. For all the methods, the samples were heated under reflux, to avoid any loss of Tc and Re. This heating step under acidic condition assures the oxidation of the Tc transferred to the liquid phase [23]. After the incineration, the sample was cooled down to room temperature and it was filtered by gravity using filter paper (Toyo 5-C; Toyo Roshi Kaisha, Ltd. Tokyo, Japan). The filter was washed with deionized water and the wash and the filtrate were transferred into a 250 ml polypropylene bottle. This solution was diluted down to 200 ml with deionized water. Then, the TEVA resin was used to concentrate the Tc and Re [24]. This extraction chromatographic resin is an aliphatic quaternary amine. Capacity factors as high as 6103 have been reported for Re. [20]. It has been used for Tc analysis because its high selectivity for Tc from environmental analysis [12,15,22,25] when Tc is loaded onto the resin as pertechnetate ion in a diluted nitric acid solution. Prepacked minicolumns (4.2 8 cm length, 0.8 mm diameter, mesh sieve 100-150 m) of this amine, which is sorbed on Amberchrom CG71, were used in these experiments. Following previous experiments developed at NIRS [26-27], the minicolumn was preconditioned with 5 ml of 8M HNO3 and 10 ml of 0.1M HNO3, using free flow. A flow rate of 1.4 ml min-1 was used to load the sample solution onto the resin. A resin washing step is necessary in order to remove the interfering elements for ICP-MS analysis (such as W for Re and Mo and Ru for Tc respectively). For 99Tc analysis in soil samples, 2M HNO3 is used [27]. However, according to those experiences previously mentioned, this nitric acid concentration would remove a great fraction of Re contained onto the resin, but Tc should be retained on it. This fact would cause a fractionation of Re and Tc atoms, which could produce an underestimation of the chemical yield. On the contrary, if 2M HNO3 were used as stripping solution, Re would be eluted, while Tc atoms would be retained onto the resin. This fact would cause an overestimation of the chemical yield. Therefore, both approaches were avoided. 20 ml of 1M HNO3 were used for washing the column. Such a process of resin washing is not a novelty; however the nitric acid concentration used is a necessary change in this chemical separation scheme, in order to avoid analyte/tracer fractionation. This washing process removes interfering elements for ICP-MS analysis without affecting the Re and Tc retained fractions in the resin [24, 26]. Both Re and Tc were stripped from the resin using 5 ml of 8M HNO3. Previous works [22,24] have shown the high efficiency of this stripping process. 9 The concentration of nitric acid is not suitable to be aspirated by the ICP-MS instrument; thus, the solution was heated to near dryness on a hot plate. The temperature was always less than 75C to avoid any Tc and Re loss in the evaporation process. Finally, the solutions were diluted to 15 ml with deionized water. An alternative to this process would be a simple sample dilution to a 0.1M HNO3 solution. However, this method would increase the analyte’s dilution factor, and, thus, this alternative would increase the limit of detection of the technique by a factor greater than 25 respecting that other method here proposed. 2.3. Apparatus. The gamma rays emitted by the 95mTc were measured on an Aloka ARC-300 NaI(Tl) scintillation counter (Tokyo, Japan) in order to calculate the Tc chemical yield. The chemical yield for Re was calculated using the isotope dilution method with the ICPMS (HP-7500, Yokogawa, Japan). Indium was used as internal standard for Mo, Ru and Tc. Operating conditions are provided on Table 2. In order to minimize the error propagation when applying the ID method, the optimum 185Re/187Re isotopic ratio was calculated according to the expression (PLEASE, CHECK REFERENCE) Ropt = (RN  RS)1/2, being Ropt = 4.21 the optimum isotopic ratio, RN the natural isotopic ratio and RS the spike isotopic ratio. For recovery calculation, the 185Re/187Re isotopic ratio determination in the analyzed solutions is required. To do that, the instrumental mass bias was determined off-line. A 16 [17] L.K.Fifield, R.S. Carling, R.G. Cresswell, P.A. Hausladen, M.L. di Tada and J.P. Day J.P. Nucl. Instr. and Meth. in Phys. Res. B 168 (2000) 427. [18] M. McCartney, V. Olive and E.M. Scott, J. Radional. Nucl. Chem. 242, (1999) 413. [19] R.C. Richter, S.R. Koirtyohann and S.S. Jurisson. J. Anal. Atom. Spectrom. 12 (1997) 557. [20] K. Tagami and S. Uchida, Anal. Chim. Acta 405 (2000) 227. [21] F. Yiou, G.M. Raisbeck, G.C. Christensen and E. Holm, J. Environ. Radioactivity 60 (2002) 61. [22] S. Uchida and K. Tagami, Anal. Chim. Acta 357 (1997) 1. [23] S. Morita, C.K. Kim, Y. Takaku, R. Seki and N. Ikeda. Appl. Radiat. Isot. 52 (1991) 531. [24] S. Uchida and K. Tagami, in Abstracts of the 9th International Symposium on Environmental Radiochemical Analysis, Maidstone UK (2002). [25] M. McCartney, K. Rajendran, V. Olive, R. Busby and P. McDonald J. Anal. Atom. Spectrom. 14 (1999) 1849. [26] K. Tagami and S. Uchida, J. Radioanal. Nucl. Chem. 239 (1999) 643. [27] S. Uchida, K. Tagami, E. Wirth and W. Rhüm. Environ. Poll. 105 (1999) 75. [28] J.L. Mas, M. García-León and J.P. Bolívar, Nucl. Instr. and Meth. in Phys. Res. A 484 (2002) 660. 17 Table captions. Table 1. Chemical yields of Tc by three different separation methods (recovery  1  counting error). a Leaching performed using 4M HNO3 after 30 minutes ashing b Leaching performed using 4M HNO3 after 90 minutes ashing c Leaching performed using 8M HNO3 after 90 minutes ashing. Table 2. Operating conditions of ICP-MS. Table 3. Chemical yields for 95mTc and Re, and Tc concentration values for the samples used in this work. a Values are determined using the chemical recovery of 95mTc. bValues are calculated using the chemical recovery of Re. Average  1 standard deviation, n = 5. 18 Sample Date of collection Method Aa Method Bb Method Cc 3014 2000/01/11 78  1 85  1 88  1 3015 2000/02/11 76  1 89  1 94  1 3016 2000/03/13 70  1 87  1 89  1 3017 2000/04/13 72  1 91  1 95  1 3021 2000/08/16 88  1 91  1 90  1 Table 1 Table Operating Conditions for ICP-MS RF power 1340 W Plasma gas flow rate 15 L/min Auxiliary gas flow rate 1.0 L/min Carrier gas flow rate 1.2 L/min Solution uptake rate 0.2 mL/min Sampling cone aperture (Ni) 1 mm Skimmer aperture (Ni) 0.4 mm Nebulizer type Babington Isotopes 95 Mo, 98 Mo, 99 Tc, 101 Ru, 102 Ru, 182 W, 185 Re, 187 Re, 115 In (Internal standard) Dwell time for 99Tc 45 ms Dwell time for Re isotopes 9 ms Sampling points per mass 3 Sweeps per replicate 1000 19 Number of replicates 4 Table 2. Sample and replicate 95m Tc recovery (%) Re recovery (%) 99 Tc a (Bq/kg d.w.) 99 Tc b (Bq/kg d.w.) 3014 76.6 ± 1.1 79.0 ± 3.3 442 ± 10 429 ± 20 97.9 ± 1.4 98.0 ± 3.3 442 ± 10 442 ± 16 96.0 ± 1.5 93.2 ± 3.6 423 ± 11 463 ± 20 93.1 ± 1.4 98.0 ± 3.9 453 ± 10 398 ± 17 87.5 ± 1.1 9 6 ± 1 1 432 ± 9 394 ± 44 3015 90.0 ± 1.1 95.6 ± 3.6 431 ± 9 406 ± 17 92.7 ± 1.4 90.3 ± 3.2 338 ± 8 369 ± 14 97.1 ± 1.5 99.5 ± 2.7 316 ± 11 303 ± 12 99.1 ± 1.5 99.1 ± 3.8 336 ± 8 336 ± 14 93.7 ± 1.2 93.3 ± 3.7 379 ± 7 381 ± 16 3016 81.4 ± 1.1 80.2 ± 2.6 456 ± 13 406 ± 17 96.1 ± 1.4 98.3 ± 4.4 474 ± 10 464 ± 22 99.0 ± 1.5 102.6 ± 3.3 459 ± 12 443 ± 17 95.8 ± 1.5 96.3 ± 3.3 462 ± 10 440 ± 16 89.3 ± 1.2 92.2 ± 3.2 465 ± 11 450 ± 18 3017 81.9 ± 1.1 84.7 ± 2.4 454 ± 27 423 ± 27 96.7 ± 1.4 99.3 ± 3.3 470 ± 10 458 ± 17 94.4 ± 1.5 92.0 ± 3.3 464 ± 10 491 ± 19 94.5 ± 1.5 97.1 ± 2.9 493 ± 12 469 ± 16 94.8 ± 1.2 91.5 ± 3.9 483 ± 9 501 ± 22 3021 98.1 ± 1.1 100.2 ± 3.0 271 ± 6 265 ± 10 98.1 ± 1.4 95.6 ± 3.7 269 ± 42 276 ± 43 98.4 ± 1.6 95.1 ± 4.1 255 ± 6 274 ± 13 94.4 ± 1.5 1 10 ± 1 4 301 ± 45 250 ± 49 90.0 ± 1.2 92.7 ± 3.6 264 ± 6 256 ± 11 Table 3 20 Figure captions. Figure 1: Fit of Tc activity concentration according to their calculations by two methods: with the recovery of 95mTc and with the recovery of Re. 21 FIGURE 1. 22