- Email: [email protected]
A new approach to determine 147Pm in irradiated fuel solutions
Talanta 78 (2009) 676–681
Contents lists available at ScienceDirect
Talanta journal homepage: www.elsevier.com/locate/talanta
A new approach to determine 147 Pm in irradiated fuel solutions René Brennetot ∗ , Guillaume Stadelmann, Céline Caussignac, Clémentine Gombert, Michèle Fouque, Christine Lamouroux Commissariat à l’Energie Atomique, Département de Physico Chimie, Service d’Etude du Comportement des Radionucléides, Laboratoire d’Analyses Nucléaires Isotopiques et Elémentaires, Bât 391, PC 33, Centre d’étude de Saclay, 91191 Gif sur Yvette cedex, France
a r t i c l e
i n f o
Article history: Received 12 September 2008 Received in revised form 5 December 2008 Accepted 11 December 2008 Available online 24 December 2008 Keywords: 147 Pm measurement ICP-MS Nuclear fuel
a b s t r a c t Developments carried out in the Laboratory of Isotopic, Nuclear and Elementary Analyses in order to quantify 147 Pm in spent nuclear fuels analyzed at the CEA within the framework of the Burn Up Credit research program for neutronic code validation are presented here. This determination is essential for safety-criticality studies. The quantity and the nature of the radionuclides in irradiated fuel solutions force us to separate the elements of interest before measuring their isotopic content by mass spectrometry. The main objective of this study is to modify the separation protocol used in our laboratory in order to recover and to measure the 147 Pm at the same time as the other lanthanides and actinides determined by mass spectrometry. A very complete study on synthetic solution (containing or not 147 Pm) was undertaken in order to determine the yield of the various stages of separation carried out before obtaining the isolated Pm fraction from the whole of the elements present in the spent fuel solutions. With the lack of natural tracer to carry out the measurement with the isotope dilution technique, the great number of isotopes in fuel, the originality of this work rests on the use of another present lanthanide in fuel to define the output of separation. The yields were measured at the conclusion of each stage of separation with two others lanthanides in order to show that one of them could be used as a tracer to correct the measurement of the 147 Pm with the separation yield. The total yield (at the conclusion of the two stages of separation) was measured at the same time by ICP-MS and liquid scintillation. This last determination made it possible to validate the use of the 147 Sm (natural) to measure the 147 Pm in ICP-MS since the outputs determined in liquid scintillation and ICP-MS (starting from the radioactive decrease of the source having been used to make the synthetic solution) were equivalent. It is the first time that such measurement is performed in ICP-MS. The measurement of the 147 Pm was finally taken on fuels UOx and MOx by using the 153 Eu like a tracer of the separation yield. The results obtained are in very good agreement with those obtained from neutronic calculation code. © 2008 Elsevier B.V. All rights reserved.
1. Introduction Promethium is a product of uranium fission. Among all Pm isotopes, measurement of isotope 147 of promethium in irradiated fuel is of prime interest (1) because this isotope contributes to the formation of samarium isotopes [1] which are retained as strong neutron absorber in Burn Up Credit studies and it is retained belong the 15 fission products of interest retained by OECD for criticality studies, (2) for calculation of fuels after power. Its measurement would make it possible to raise uncertainties concerning a possible undervaluation of the fission yields.
∗ Corresponding author. Tel.: +33 169 08 56 23; fax: +33 169 08 54 11. E-mail address: [email protected] (R. Brennetot). 0039-9140/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.talanta.2008.12.022
Different methods are reported in the literature for 147 Pm separation and measurement (principally by liquid scintillation counting) in various matrices such as urine, environmental samples and nuclear fuel [2–5]. For urine samples, Eichuk describes a method using dynamic ion exchange chromatography to separate 147 Pm from urine matrix and a measurement of the Pm fraction by liquid scintillation counting. For environmental samples, Yoshida et al. reported a rapid separation method using HPLC with HIBA (Alpha-hydroxyisobutyric acid) stepwise eluent method for separation of lanthanides Sm, Pm, Eu, Nd with liquid scintillation counting for 147 Pm and 151 Sm determination and inductively coupled plasma mass spectrometry for Eu, Nd and Sm measurements. Jerome used a co precipitation method for Pm determination. The lanthanides are first co precipitated and then separated by cation exchange chromatography in HIBA medium. 147 Pm measurement is then performed by liquid scintillation counting. This procedure has been
R. Brennetot et al. / Talanta 78 (2009) 676–681
used for various matrices such as sediments, power station effluent discharges, silt, . . . G Seeber et al synthesized a new coated silica material for separation of radioactive lanthanides 147 Pm and 152 Eu, both measured by liquid scintillation counting. In irradiated fuel matrix, only one method has been previously reported and developed by Adriansen [6]. They first separate 147 Pm from other lanthanides using the Ln resin and a 146 Nd spike to obtain the separation yield. 147 Pm measurement is performed with liquid scintillation counting. Isotopic and elementary measurements of radionuclides in spent nuclear fuel samples are principally done with mass spectrometry techniques including thermal ionisation mass spectrometry (TIMS) [7,8], multi collector inductively coupled plasma mass spectrometry (MC-ICP-MS) [9–11], sector field inductively coupled plasma mass spectrometry (SF-ICP-MS) [12] and quadrupole inductively coupled plasma mass spectrometry (Q-ICPMS) [13–18]. In all cases, when separation steps are needed for isotopic content determination, isotopic dilution technique is used with double spike in order to be independent of separation yields. Several works compared the ICP-MS with radiometric techniques like alpha spectrometry or gamma spectrometry for radionuclide measurements [19–21]. The goal of this work is to obtain a new protocol allowing the measurement of 147 Pm by quadripolar ICP-MS without adding stages. Protocols available in our laboratory allow the separation of the various fission products as well as the determination of their concentration and isotopic content by mass spectrometry techniques. Beta-decay of 147 Pm leads to formation of isobaric 147 Sm. The resolution necessary to separate these two peaks (735,000), is higher than the resolution usually accessible by mass spectrometry. Thus, the determination of the 147 Pm by mass spectrometry can consequently be done, only after chemical separation of these two elements. The goal of this work is to adapt the available protocols, in order to be able to measure 147 Pm. Initially, fission products (FP) are separated from uranium and plutonium, which are the two major components of irradiated fuels, using a previously described method involving ion exchange resin [6,7]. The fission products collected fraction FP which contains elements such as samarium and promethium is then injected in HPLC in order to separate the various components. Finally, the Pm fraction obtained from HPLC will be analyzed by ICP-MS by external calibration. It is obvious that we need to follow exactly the comportment of Pm during the separation steps in order to correct the ICP-MS measurement with the separation yield. The originality of this work rests on the use of another lanthanide isotope present in fuel, thus making it possible to define the yield of all the separation steps. In this work, different lanthanides such as Eu and Gd were studied to verify if they can be used as a tracer of the separation steps (Fig. 1). The resin classically used to carry out the first U/Pu/FP separation is an anion resin AG1X4 (Biorad) in nitric medium. However, and as clearly explained by Marhol [22], lanthanides do not have the same distribution coefficients for a given acidity on AG1X4 resin in nitric medium precluding therefore the use of these conditions. The lanthanides interaction of an anion resin AG1X10 seems similar in hydrochloric medium according to the literature [3]. Consequently, the output of the first separation on resin AG1X10 in hydrochloric medium was studied in order to determine if europium and gadolinium had the same retention under these conditions according to the strategy presented in Fig. 1. The final goal of this work is to develop a new protocol for 147 Pm measurement in nuclear fuel samples (using the actual protocol for separation yield measurement at the same time). A complete study on synthetic solution (containing or not Pm) was undertaken in order to determine the yield of the various separation steps carried out before obtaining the pure Pm fraction. This
677
Fig. 1. Synoptic of this study.
study shows that the separation yield can be followed by the use of another lanthanide used as a tracer of the separation steps. The total yield (at the conclusion of the two separation stages) was measured at the same time by ICP-MS and liquid scintillation counting. Finally, the results obtained for the 147 Pm measurement in two types of fuel with high burn up (70 GWd/t) UOx and MOx type are shown. 2. Experimental All the concentrations reported in this paper are expressed in ng g−1 . Activity concentrations measured by liquid scintillation counting are expressed in Bq g−1 . 2.1. Reagents and standards All dilutions have been performed in nitric acid at 2% (v/v). This acid is obtained by dilution of nitric acid (65% Normatom Prolabo) in deionised water (conductivity 18.2 M) obtained from a Milli Q system (Millipore). Sub-boiling nitric acid is manufactured at the laboratory starting from 65% HNO3 (Normatom Prolabo). Hydrochloric acid (33–36% Optima Prolabo) is used for gravimetric separations and ICP-MS experiments. A multi-element Tuning Solution (Spex) at 1 g/L was used every day for the instrument optimization and short-term stability tests. All calibration curves have been performed using SPEX solutions (1 mg/L) of samarium, gadolinium and europium and uranium (10 mg/L) for matrix matching. A solution of 147 Pm certified in activity was obtained from CERCA (LEA, AREVA). The initial activity was 800 ± 16 kBq/g. This solution has been used for experiments performed with fuel simulated solutions and calibration of the liquid scintillation counting. 115 In (SPEX) is used as internal standard for ICP-MS measurements. 2-Hydroxy-methylbutyric acid (HMB) is used as eluent for HPLC lanthanide separation (Sigma–Aldrich). pH of the mobile phase was adjusted with a 25% ammonia solution (Normatom, Prolabo). Arsenazo III (Sigma–Aldrich) is used as a post-column reactant for samarium detection with UV detector.
678
R. Brennetot et al. / Talanta 78 (2009) 676–681
2.2. Ion exchange resins
Table 1 ICP-MS operating parameters.
An anion exchange Dowex AG1X10 (Bio Rad Laboratories) resin, 50–100 mesh, was used for U/Pu/FP separation. For conditioning, the dry resin is placed in an eco column provided by Eichrom, the height of resin must be of 2.5 cm ± 1 mm in the column. The resin is then washed with concentrated hydrochloric acid (10 M). 15 mL of the same acid is sufficient to wash and condition the resin. The selected protocol is the one already used at the laboratory for the separations carried out on resin in hydrochloric medium. Aliquot of solution is taken and dried in a Teflon beaker. 1 mL of OPTIMA HCl 33–36% is used to rinse the beaker, this aliquot is deposited on the head column. 4 mL of HCl 10 M is necessary to elute completely the FP fraction containing the 147 Pm; under these conditions U and Pu are trapped on the resin. Only the first fraction containing the fission products was recovered in a container out of Teflon and weighed precisely. This fraction, in 10 M HCl medium, was diluted in 0.5 M HNO3 for measurement by ICP-MS.
Forward rf power Reflected rf power Coolant gas flow rate Nebulizer gas flow rate Auxilliary gas flow rate Nebulizer type Spray chamber Detection mode Dwell time Sweeps Sampling and skimmer cones
2.3. Samples Simulated solutions were used for all separation yield determination of lanthanides. They are prepared by weight of SPEX standards solutions and contain only U, Sm, Eu, Gd and Pm for yield studies including HPLC separation. The spent fuel samples were UOx and MOx types. They were dissolved in hot cells at CEA Marcoule in the ATALANTE facility. After complete dissolution, the samples are diluted and sent to CEA Saclay for elementary and isotopic analysis. 2.4. Analytical techniques 2.4.1. ICP-MS The experiments were carried out on a quadrupole ICP-MS X series (Thermo Fischer Scientific). This ICP-MS was the first X7 modified in order to work with radioactive materials as previously described [23]. Sample introduction to the plasma was made via a quartz concentric nebulizer (0.4 mL/min) and a quartz bead impact spray chamber. A CETAC ASX 260 auto-sampler was used in order to minimize exposure to radioactive materials. All of the experimental parameters are daily optimized with a Test solution containing several elements in order to obtain the maximum count rates on 115 In and 238 U for this tuning. After acquisition of a short-term stability test in standard mode (sensitivity, stability and oxide level test), the instrument is ready to perform measurements by external calibrations.
1400 W <2 W 15 L min−1 0.95 L min−1 0.8 L min−1 Quartz concentric Bead impact cooled with peltier effect at 3 ◦ C Pulse counting 10 ms 300 Ni
Experimental parameters for mass spectrometric measurements are summarized in Table 1. The external calibration was carried out with multi elementary standards prepared with SPEX solutions. Indium was selected as internal standard for measurement in ICP-MS because it is the only one which can be used for the analysis of fuel. The dilution medium, 0.5 M HNO3 is selected in order to make sure that the standards, the simulated solution and FP fractions can be measured using the same calibration (absence of matrix effect). 2.4.2. HPLC The chromatographic system used consists of several modules. A principal chromatographic pump (Merck-Hitachi) delivering the mobile phase made up of HMB in gradient mode (in quantity and pH) with a flow of 0.8 mL/min. A Rhéodyne valve six ways in peek is used with a 100 L injection loop in peek. The column used is a cation exchange column NUCLEOSIL SA of length and internal diameter respectively 250 and 4.6 mm, thermostated at 25 ◦ C. A secondary pump (Gilson) is used for delivery of arsenazo III necessary for the UV/vis detection of Sm at 658 nm (Jasco). A radiometric detector (PerkinElmer) is used for radio elements detection. 2.4.3. Liquid scintillation counting Measurement of 147 Pm by liquid scintillation counting were performed with two instruments: TriCarb 2900TR and Wallac Oy 1414 Guardian. The dilutant used is selected to be close to the final medium which will be used to count on the solutions resulting from the HPLC and analyzed by ICP-MS. 3. Results and discussion The measurement of 147 Pm suffers from an isobaric interference with 147 Sm implying therefore separations steps. The main objective of this work is to find a tracer of the separation steps in order to correct the obtained value for Pm concentration in ICP-MS of the
Fig. 2. Mass spectrum of spent fuel from 130 uma to 160 u.
R. Brennetot et al. / Talanta 78 (2009) 676–681 Table 2 Yield obtained for Gd and Eu for an aliquot of 1 mL of synthetic solution after gravimetric separation by AG1X10 resin in HCl medium.
Aliquot 1 Aliquot 2 Aliquot 3
Eu yield
Gd yield
99 ± 2 99 ± 2 98 ± 2
99 ± 2 99 ± 2 98 ± 2
separation yield to be able to measure 147 Pm in spent fuels taking into account all the separation steps to correct the final measurement. It was decided to study the separation yield at each step of the analytical process used for 147 Pm measurement in order to find a marker of the separation to correct the final result of the separation yield. (1) Gravitational separation on resin: FP, U, Pu. (2) HPLC separation of actinides and lanthanides. (3) Measurement of the 147 Pm fraction free from 147 Sm by ICP-MS with a calibration curve at mass 147 with natural samarium. Ideally a double spike with the use of the isotope dilution technique would be optimal to overcome the measurement of the separation yields, but such a tracer is not available. It could be interesting to use another isotope of Pm as the 148 Pm but its lifetime is too much short to be used. Consequently, we decided to use a chemical analogue which would have the same chemical behaviour as Pm. The irradiated fuel is a complex solution which comprises many elements including lanthanides like shown in Fig. 2. Consequently, it is possible to use other lanthanides after ensuring that they have the same chemical behaviour as Pm with respect to the separation stages considered. Europium and gadolinium have been tested with respect to the separation stages (gravimetric separation + HPLC) in order to check their potential as tracer of the separation. We have chosen these two lanthanides for their proximity in mass and because they can be measured with a very small uncertainty, using the isotope dilution technique. This technique, well described in the literature [24] allows the determination of isotopic ratios and radionuclide concentrations with a very small uncertainty. 3.1. Gravimetric separation In order to analyze and to determine the content of 147 Pm in a spent nuclear fuel, it is necessary to separate the 147 Pm from the other elements such as fission products, U and Pu. This separation is carried out in two stages: the first separation is performed with an ion exchange resin in order to separate rapidly FP from the U and Pu matrix. Then, FP are separated by HPLC. In order to minimize the dose rate received by the experimenters, the yield of the first separation step for two selected lanthanides (Eu and Gd) was measured using a synthetic solution including only uranium, gadolinium and europium (natural) in proportions similar to those of spent nuclear fuels. This synthetic solution contains around 680 g/g U, 210 ng/g Eu and 390 ng/g Gd. Three tests were carried out. An aliquot of 1 mL of this synthetic solution is used, in order to make sure that a sufficient quantity of matter is able at the end of the first separation stages, to determine the yield by ICP-MS. Eu, and Gd concentrations are perfectly known as the synthetic solution is made by weighting. Concentrations in Eu and Gd were measured before and after gravimetric separation in order to determine the yield of separation and to prove that it could be considered as equivalent. The yields were measured for isotopes 151 and 153 of europium and 157 and 158 of gadolinium. The results from these three separations on resin are presented in Table 2. Error bars correspond to
679
reproducibility observed (4 calibrations ICP-MS with each time two test specimens of the fraction). At this stage, it can be considered that europium and gadolinium could be used like a tracer of the first separation stage since the separation yields are equivalent, very reproducible and close to 100%, suggesting therefore that no fractionation is induced by the gravimetric separation with the protocol chosen here. 3.2. HPLC A new synthetic solution containing 147 Pm at the concentration of 9 ng/g has been prepared. The 5 mL solution thus recovered at the end of the first separation previously described is put in a Teflon beaker and dried. The residue is then reconditioned in nitric medium with 5 mL of 0.2N Sub-boiling nitric acid compatible with HPLC. It is then injected on a HPLC chain confined in a glove box to separate Sm/Pm/Eu/Gd. The separation yield for Pm is obtained by calculation of the true concentration to date. The concentration in 147 Pm is measured by ICP-MS by using the 147 Sm as standard (it is supposed that Sm and Pm have the same response in ICP-MS). The results obtained on the three tests carried out in glove box are presented in Table 3. To ensure that the response of Pm and Sm could be expected equivalent in ICP-MS, the yield of separation for Pm is measured in the same solution by liquid scintillation counting, an independent measurement technique without any interference since the only beta emitter in the synthetic solution is 147 Pm. The results presented in Table 3 show that the same yields are obtained by liquid scintillation counting and by ICP-MS with samarium 147 used as a standard for external calibration for Pm measurement. This indicates at this stage, that europium and gadolinium could be used as tracer of the whole separation steps envisaged. 3.3. Analysis of spent fuels Previous developments with synthetic solutions show that the yield at the end of the separation stages with selected conditions are equivalent for selected lanthanides Pm, Eu, Gd. Thus 153 Eu and 157 Gd could be used as tracers of the separation yield. In addition it has been shown that 147 Sm could be used to carry out the range of calibration to measure 147 Pm. Two spent fuel solutions were analyzed. Separations were carried out on the MOx and UOx fuels. Two aliquots of 250 l for each fuel were taken for separation steps. Chromatogram of one of the aliquot of UOx sample is presented in Fig. 3. As shown in Fig. 2, there is lots of peaks in the lanthanide zone for a spent fuel what results in to create interferences which were not taken into account up to now. Finally 153 Eu was chosen as the tracer of the separation, because of the presence of some oxides (BaO, CeO, PrO) that induce an over estimation of the absolute content of gadolinium in the fuel without separation step. For this purpose, it has been experimentally verified that caesium do not produce a sufficient oxide quantity to interfere at mass 149, Table 3 Yield obtained for Gd, Eu and Pm for an aliquot of 1 mL of synthetic solution after gravimetric separation and HPLC. The first column for Pm is the yield determined by ICPMS with Sm external calibration, the second one is the determination by liquid scintillation counting.
Aliquot 4 Aliquot 5 Aliquot 6
Eu yield
Gd yield
Pm yield
81 ± 3 71 ± 3 81 ± 3
80 ± 3 73 ± 3 83 ± 3
81 ± 3 70 ± 3 77 ± 3
79 ± 3 70 ± 3 78 ± 3
680
R. Brennetot et al. / Talanta 78 (2009) 676–681
returned by isotopic dilution with a variation observed lower than 1% for UOx sample and less than 3% for MOx sample. The 147 Pm concentration obtained were then compared with prediction calculation according to DARWIN 2.2 code with JEFF 2.2 library. Less than a 10% variation is obtained between the measurement and the result obtained starting from the computer code. However this variation is covered by the uncertainty of measurement which validates this protocol. 4. Conclusion
Fig. 3. Chromatogram obtained with radiometric detection for separation of the FP fraction obtained at the first gravimetric separation of UOx GGU1 fuel.
150, 151, 153 when barium, praseodymium and cerium produces a large amount of oxides that interferes all the isotopes of gadolinium. The results are presented in Table 4. Uncertainties were given on several levels in order to try to take into account all stages leading to the final result. The analytical uncertainty given in ICP-MS takes into account all the analytical stages, of the preparation of the sample received at the laboratory with its analysis: preparation of the standards by weighing, uncertainty on the concentrations of the standards (calculated starting from the certificates of analysis of the standard solutions), uncertainty on the calibration, stability of the apparatus, estimate of the accuracy of the analyzes starting from standards whose concentration is known, variation of measurements on two distinct test specimens, derives possible from the apparatus over the duration of measurements. Uncertainties on the concentrations determined in ICP-MS were given starting from LANI, homemade software which integrates the sources of uncertainties such as dilutions, uncertainty on the standards and the hyperboles of confidence of the calibration line. The values of concentrations in 153 Eu obtained before and after separations are compiled in order to obtain the output of the stages of separation. Resulting uncertainty is given with the method described by Kragten [25]. Combined uncertainty is calculated starting from the law of propagation of uncertainties like defined in EURACHEM/CITAC guide [26]. The result obtained on the absolute concentration of 153 Eu was first of all compared with the concentration returned by isotopic dilution (ID) in June 2006 (taken as reference value). As can be seen in Table 4, europium concentration is coherent with the result Table 4 Results obtained for 153 Eu and 147 Pm concentration in spent fuels. Separation yields applied for 147 Pm concentration are calculated from 153 Eu concentration measured by Q-ICP-MS before and after separation steps. 153
Eu (ng/g sol)
153
Eu (ng/g sol) reference value
147
Pm (ng/g sol)
UOx Before sep steps: 161.7 After sep steps: 128.3
163.3 ± 3.2
27.5 ± 2.7
MOx Before sep steps: 266.4 After sep steps: 217.5
273.2 ± 5.9
40.14 ± 4.06
With an aim of determining the concentration in 147 Pm in UOx and MOx spent fuel solutions, a durable method was developed in order to validate the separation stages necessary to this measurement (exchange resin and HPLC). This new measurement was optimized to complete a method previously used for lanthanides determination by isotope dilution technique. These tests made it possible to show that lanthanides Eu, Gd and Pm had a similar behaviour with respect to the selected resin (AG1X10) in hydrochloric medium in this type of matrix. We have shown that Eu and Gd had the same behaviour during all the separations steps, making it possible their use as a tracer of the separation for Pm measurement. Checking of the yield at the conclusion of the stages of separation by liquid scintillation made it possible to validate on the one hand the yields determined by ICP-MS, but also to validate the use of the natural 147 Sm to measure the 147 Pm after separation. The yield thus determined for the 153 Eu could be applied to the measurement of the 147 Pm in fuel. These results also were validated by ICP-MS with a difference lower than 2% compared to the results resulting from the determinations by isotopic dilution in May and October 2006 to the measurements of absolute concentrations of the isotope 153 Eu used for the follow-up and the determination of the yield on the totality of the experiment. The measurements obtained in ICP-MS for the 147 Pm and measurements of output of 153 Eu made it possible to calculate the concentrations in 147 Pm in fuel. These results are completely coherent with the values returned by calculation code with a difference less than 10% and totally covered by uncertainty measurement. Acknowledgments The authors wish to thank Cecile Riffard (CEA Cadarache, DER/SPRC/LECy) for its support for calculation codes. Authors wish to thank also Carole frechou (DEN/DANS/DPC/SECR/LANIE) for her advices and Pascal Fichet (DEN/DANS/DPC/SECR/LANIE) for his uncertainty determination program. References [1] B. Roque, N. Thiollay, P. Marimbeau, A. Barreau, A. Tsilanizara, C. Garzenne, F. Marcel, H. Toubon, C. Garats, Park, MG, (Ed.), Proceedings PHYSOR-2002, Seoul, Korea, 7–10 October 2002. [2] S.M. Jerome, Sci. Total Environ. 70 (1988) 275. [3] M. Yoshida, S.H. Sumiya, H. Watanabe, K. Tobita, J. Radioanal. Nucl. Chem. 197 (1995) 219. [4] G. Seeber, P. brunner, M.R. Buchmeiser, G.K. Bonn, J. Chromatogr. A 848 (1999) 193. [5] S. Eichuk, C.A. Lucy, K.I. Burns, Anal. Chem. 64 (1992) 2339. [6] L. Adriaensen, M. Gysemans, C. Hurtgen, D. Boulanger, ENC 2007 Conference, Brussels, September 16–20, 2007. [7] F. Chartier, M. Aubert, M. Salmon, M. Tabarant, B.H. Tran, J. Anal. Atom. Spectrom. 14 (1999) 1661. [8] F. Chartier, M. Aubert, M. Pilier, Fresenius J. Anal. Chem. 364 (1999) 320. [9] H. Isnard, R. Brennetot, C. Caussignac, N. Caussignac, F. Chartier, Int. J. Mass Spectrom. 246 (2005) 66. [10] I. Gunther Leopold, N. Kivel, J. Kobler, B. Wernli, Anal. Bioanal. Chem. (2008) 503. [11] H. Isnard, M. Aubert, P. Blanchet, R. Brennetot, F. Chartier, V. Geertsen, F. Mannuguerra, Spectrochim. Acta 61B (2006) 150.
R. Brennetot et al. / Talanta 78 (2009) 676–681 [12] J.S. Becker, W. Kerl, H.J. Dietze, Anal. Chim. Acta 387 (1999) 145. [13] D. Solatie, P. Carbol, M. Betti, F. Bocci, T. Hiernaut, V.V. Rondinella, J. Cobos, Fresenius J. Anal. Chem. 368 (2000) 88. [14] C.S. Kim, C.K. Kim, P. Martin, U. Sansone, J. Anal. Atom. Spectrom. 22 (2007) 827. [15] J.I. Garcia Alonso, F. Sena, P. Arbore, M. Betti, L. Koch, J. Anal. Atom. Spectrom. 10 (1995) 381. [16] J. Comte, P. Bienvenu, E. Brochard, J.M. Fernandez, G. Andreoletti, J. Anal. Atom. Spectrom. 18 (2003) 702. [17] J.I. Garcia Alonso, J.F. Babelot, J.P. Glatz, O. Cromboom, L. Koch, Radiochim. Acta 62 (1993) 71. [18] J.S. Becker, R.S. Soman, K.L. Sutton, J.A. Caruso, H.J. Dietze, J. Anal. Atom. Spectrom. 14 (1999) 933. [19] S.F. Boulyga, C. Testa, D. Desideri, J.S. Becker, J. Anal. Atom. Spectrom. 10 (1995) 381.
681
[20] N. Baglan, C. Bouvier Capely, C. Cossonnet, Radiochim. Acta 90 (2002) 267. [21] P. Roos, H. Xiaolin, Anal. Chim. Acta 608 (2008) 105. [22] M. Marhol, Ion Exchangers in Analytical Chemistry. Their Properties and Use in Inorganic Chemistry, vol. XIV, Elsevier Scientific Publishing Company, 1982, 585 p. [23] G. Favre, R. Brennetot, F. Chartier, P. Vitorge, Int. J. Mass Spectrom. 265 (2007) 15. [24] K.G. Heumann, Int. J. Mass Spectrom. 118/119 (1995) 575. [25] J. Kragten, Analyst 119 (1994) 2161. [26] EURACHEM Secretariat, EURACHEM/CITAC Guide Quantifying Uncertainty in Analytical Measurement, second ed., EURACHEM Secretariat, BAM, Berlin, 2000.
Contents lists available at ScienceDirect
Talanta journal homepage: www.elsevier.com/locate/talanta
A new approach to determine 147 Pm in irradiated fuel solutions René Brennetot ∗ , Guillaume Stadelmann, Céline Caussignac, Clémentine Gombert, Michèle Fouque, Christine Lamouroux Commissariat à l’Energie Atomique, Département de Physico Chimie, Service d’Etude du Comportement des Radionucléides, Laboratoire d’Analyses Nucléaires Isotopiques et Elémentaires, Bât 391, PC 33, Centre d’étude de Saclay, 91191 Gif sur Yvette cedex, France
a r t i c l e
i n f o
Article history: Received 12 September 2008 Received in revised form 5 December 2008 Accepted 11 December 2008 Available online 24 December 2008 Keywords: 147 Pm measurement ICP-MS Nuclear fuel
a b s t r a c t Developments carried out in the Laboratory of Isotopic, Nuclear and Elementary Analyses in order to quantify 147 Pm in spent nuclear fuels analyzed at the CEA within the framework of the Burn Up Credit research program for neutronic code validation are presented here. This determination is essential for safety-criticality studies. The quantity and the nature of the radionuclides in irradiated fuel solutions force us to separate the elements of interest before measuring their isotopic content by mass spectrometry. The main objective of this study is to modify the separation protocol used in our laboratory in order to recover and to measure the 147 Pm at the same time as the other lanthanides and actinides determined by mass spectrometry. A very complete study on synthetic solution (containing or not 147 Pm) was undertaken in order to determine the yield of the various stages of separation carried out before obtaining the isolated Pm fraction from the whole of the elements present in the spent fuel solutions. With the lack of natural tracer to carry out the measurement with the isotope dilution technique, the great number of isotopes in fuel, the originality of this work rests on the use of another present lanthanide in fuel to define the output of separation. The yields were measured at the conclusion of each stage of separation with two others lanthanides in order to show that one of them could be used as a tracer to correct the measurement of the 147 Pm with the separation yield. The total yield (at the conclusion of the two stages of separation) was measured at the same time by ICP-MS and liquid scintillation. This last determination made it possible to validate the use of the 147 Sm (natural) to measure the 147 Pm in ICP-MS since the outputs determined in liquid scintillation and ICP-MS (starting from the radioactive decrease of the source having been used to make the synthetic solution) were equivalent. It is the first time that such measurement is performed in ICP-MS. The measurement of the 147 Pm was finally taken on fuels UOx and MOx by using the 153 Eu like a tracer of the separation yield. The results obtained are in very good agreement with those obtained from neutronic calculation code. © 2008 Elsevier B.V. All rights reserved.
1. Introduction Promethium is a product of uranium fission. Among all Pm isotopes, measurement of isotope 147 of promethium in irradiated fuel is of prime interest (1) because this isotope contributes to the formation of samarium isotopes [1] which are retained as strong neutron absorber in Burn Up Credit studies and it is retained belong the 15 fission products of interest retained by OECD for criticality studies, (2) for calculation of fuels after power. Its measurement would make it possible to raise uncertainties concerning a possible undervaluation of the fission yields.
∗ Corresponding author. Tel.: +33 169 08 56 23; fax: +33 169 08 54 11. E-mail address: [email protected] (R. Brennetot). 0039-9140/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.talanta.2008.12.022
Different methods are reported in the literature for 147 Pm separation and measurement (principally by liquid scintillation counting) in various matrices such as urine, environmental samples and nuclear fuel [2–5]. For urine samples, Eichuk describes a method using dynamic ion exchange chromatography to separate 147 Pm from urine matrix and a measurement of the Pm fraction by liquid scintillation counting. For environmental samples, Yoshida et al. reported a rapid separation method using HPLC with HIBA (Alpha-hydroxyisobutyric acid) stepwise eluent method for separation of lanthanides Sm, Pm, Eu, Nd with liquid scintillation counting for 147 Pm and 151 Sm determination and inductively coupled plasma mass spectrometry for Eu, Nd and Sm measurements. Jerome used a co precipitation method for Pm determination. The lanthanides are first co precipitated and then separated by cation exchange chromatography in HIBA medium. 147 Pm measurement is then performed by liquid scintillation counting. This procedure has been
R. Brennetot et al. / Talanta 78 (2009) 676–681
used for various matrices such as sediments, power station effluent discharges, silt, . . . G Seeber et al synthesized a new coated silica material for separation of radioactive lanthanides 147 Pm and 152 Eu, both measured by liquid scintillation counting. In irradiated fuel matrix, only one method has been previously reported and developed by Adriansen [6]. They first separate 147 Pm from other lanthanides using the Ln resin and a 146 Nd spike to obtain the separation yield. 147 Pm measurement is performed with liquid scintillation counting. Isotopic and elementary measurements of radionuclides in spent nuclear fuel samples are principally done with mass spectrometry techniques including thermal ionisation mass spectrometry (TIMS) [7,8], multi collector inductively coupled plasma mass spectrometry (MC-ICP-MS) [9–11], sector field inductively coupled plasma mass spectrometry (SF-ICP-MS) [12] and quadrupole inductively coupled plasma mass spectrometry (Q-ICPMS) [13–18]. In all cases, when separation steps are needed for isotopic content determination, isotopic dilution technique is used with double spike in order to be independent of separation yields. Several works compared the ICP-MS with radiometric techniques like alpha spectrometry or gamma spectrometry for radionuclide measurements [19–21]. The goal of this work is to obtain a new protocol allowing the measurement of 147 Pm by quadripolar ICP-MS without adding stages. Protocols available in our laboratory allow the separation of the various fission products as well as the determination of their concentration and isotopic content by mass spectrometry techniques. Beta-decay of 147 Pm leads to formation of isobaric 147 Sm. The resolution necessary to separate these two peaks (735,000), is higher than the resolution usually accessible by mass spectrometry. Thus, the determination of the 147 Pm by mass spectrometry can consequently be done, only after chemical separation of these two elements. The goal of this work is to adapt the available protocols, in order to be able to measure 147 Pm. Initially, fission products (FP) are separated from uranium and plutonium, which are the two major components of irradiated fuels, using a previously described method involving ion exchange resin [6,7]. The fission products collected fraction FP which contains elements such as samarium and promethium is then injected in HPLC in order to separate the various components. Finally, the Pm fraction obtained from HPLC will be analyzed by ICP-MS by external calibration. It is obvious that we need to follow exactly the comportment of Pm during the separation steps in order to correct the ICP-MS measurement with the separation yield. The originality of this work rests on the use of another lanthanide isotope present in fuel, thus making it possible to define the yield of all the separation steps. In this work, different lanthanides such as Eu and Gd were studied to verify if they can be used as a tracer of the separation steps (Fig. 1). The resin classically used to carry out the first U/Pu/FP separation is an anion resin AG1X4 (Biorad) in nitric medium. However, and as clearly explained by Marhol [22], lanthanides do not have the same distribution coefficients for a given acidity on AG1X4 resin in nitric medium precluding therefore the use of these conditions. The lanthanides interaction of an anion resin AG1X10 seems similar in hydrochloric medium according to the literature [3]. Consequently, the output of the first separation on resin AG1X10 in hydrochloric medium was studied in order to determine if europium and gadolinium had the same retention under these conditions according to the strategy presented in Fig. 1. The final goal of this work is to develop a new protocol for 147 Pm measurement in nuclear fuel samples (using the actual protocol for separation yield measurement at the same time). A complete study on synthetic solution (containing or not Pm) was undertaken in order to determine the yield of the various separation steps carried out before obtaining the pure Pm fraction. This
677
Fig. 1. Synoptic of this study.
study shows that the separation yield can be followed by the use of another lanthanide used as a tracer of the separation steps. The total yield (at the conclusion of the two separation stages) was measured at the same time by ICP-MS and liquid scintillation counting. Finally, the results obtained for the 147 Pm measurement in two types of fuel with high burn up (70 GWd/t) UOx and MOx type are shown. 2. Experimental All the concentrations reported in this paper are expressed in ng g−1 . Activity concentrations measured by liquid scintillation counting are expressed in Bq g−1 . 2.1. Reagents and standards All dilutions have been performed in nitric acid at 2% (v/v). This acid is obtained by dilution of nitric acid (65% Normatom Prolabo) in deionised water (conductivity 18.2 M) obtained from a Milli Q system (Millipore). Sub-boiling nitric acid is manufactured at the laboratory starting from 65% HNO3 (Normatom Prolabo). Hydrochloric acid (33–36% Optima Prolabo) is used for gravimetric separations and ICP-MS experiments. A multi-element Tuning Solution (Spex) at 1 g/L was used every day for the instrument optimization and short-term stability tests. All calibration curves have been performed using SPEX solutions (1 mg/L) of samarium, gadolinium and europium and uranium (10 mg/L) for matrix matching. A solution of 147 Pm certified in activity was obtained from CERCA (LEA, AREVA). The initial activity was 800 ± 16 kBq/g. This solution has been used for experiments performed with fuel simulated solutions and calibration of the liquid scintillation counting. 115 In (SPEX) is used as internal standard for ICP-MS measurements. 2-Hydroxy-methylbutyric acid (HMB) is used as eluent for HPLC lanthanide separation (Sigma–Aldrich). pH of the mobile phase was adjusted with a 25% ammonia solution (Normatom, Prolabo). Arsenazo III (Sigma–Aldrich) is used as a post-column reactant for samarium detection with UV detector.
678
R. Brennetot et al. / Talanta 78 (2009) 676–681
2.2. Ion exchange resins
Table 1 ICP-MS operating parameters.
An anion exchange Dowex AG1X10 (Bio Rad Laboratories) resin, 50–100 mesh, was used for U/Pu/FP separation. For conditioning, the dry resin is placed in an eco column provided by Eichrom, the height of resin must be of 2.5 cm ± 1 mm in the column. The resin is then washed with concentrated hydrochloric acid (10 M). 15 mL of the same acid is sufficient to wash and condition the resin. The selected protocol is the one already used at the laboratory for the separations carried out on resin in hydrochloric medium. Aliquot of solution is taken and dried in a Teflon beaker. 1 mL of OPTIMA HCl 33–36% is used to rinse the beaker, this aliquot is deposited on the head column. 4 mL of HCl 10 M is necessary to elute completely the FP fraction containing the 147 Pm; under these conditions U and Pu are trapped on the resin. Only the first fraction containing the fission products was recovered in a container out of Teflon and weighed precisely. This fraction, in 10 M HCl medium, was diluted in 0.5 M HNO3 for measurement by ICP-MS.
Forward rf power Reflected rf power Coolant gas flow rate Nebulizer gas flow rate Auxilliary gas flow rate Nebulizer type Spray chamber Detection mode Dwell time Sweeps Sampling and skimmer cones
2.3. Samples Simulated solutions were used for all separation yield determination of lanthanides. They are prepared by weight of SPEX standards solutions and contain only U, Sm, Eu, Gd and Pm for yield studies including HPLC separation. The spent fuel samples were UOx and MOx types. They were dissolved in hot cells at CEA Marcoule in the ATALANTE facility. After complete dissolution, the samples are diluted and sent to CEA Saclay for elementary and isotopic analysis. 2.4. Analytical techniques 2.4.1. ICP-MS The experiments were carried out on a quadrupole ICP-MS X series (Thermo Fischer Scientific). This ICP-MS was the first X7 modified in order to work with radioactive materials as previously described [23]. Sample introduction to the plasma was made via a quartz concentric nebulizer (0.4 mL/min) and a quartz bead impact spray chamber. A CETAC ASX 260 auto-sampler was used in order to minimize exposure to radioactive materials. All of the experimental parameters are daily optimized with a Test solution containing several elements in order to obtain the maximum count rates on 115 In and 238 U for this tuning. After acquisition of a short-term stability test in standard mode (sensitivity, stability and oxide level test), the instrument is ready to perform measurements by external calibrations.
1400 W <2 W 15 L min−1 0.95 L min−1 0.8 L min−1 Quartz concentric Bead impact cooled with peltier effect at 3 ◦ C Pulse counting 10 ms 300 Ni
Experimental parameters for mass spectrometric measurements are summarized in Table 1. The external calibration was carried out with multi elementary standards prepared with SPEX solutions. Indium was selected as internal standard for measurement in ICP-MS because it is the only one which can be used for the analysis of fuel. The dilution medium, 0.5 M HNO3 is selected in order to make sure that the standards, the simulated solution and FP fractions can be measured using the same calibration (absence of matrix effect). 2.4.2. HPLC The chromatographic system used consists of several modules. A principal chromatographic pump (Merck-Hitachi) delivering the mobile phase made up of HMB in gradient mode (in quantity and pH) with a flow of 0.8 mL/min. A Rhéodyne valve six ways in peek is used with a 100 L injection loop in peek. The column used is a cation exchange column NUCLEOSIL SA of length and internal diameter respectively 250 and 4.6 mm, thermostated at 25 ◦ C. A secondary pump (Gilson) is used for delivery of arsenazo III necessary for the UV/vis detection of Sm at 658 nm (Jasco). A radiometric detector (PerkinElmer) is used for radio elements detection. 2.4.3. Liquid scintillation counting Measurement of 147 Pm by liquid scintillation counting were performed with two instruments: TriCarb 2900TR and Wallac Oy 1414 Guardian. The dilutant used is selected to be close to the final medium which will be used to count on the solutions resulting from the HPLC and analyzed by ICP-MS. 3. Results and discussion The measurement of 147 Pm suffers from an isobaric interference with 147 Sm implying therefore separations steps. The main objective of this work is to find a tracer of the separation steps in order to correct the obtained value for Pm concentration in ICP-MS of the
Fig. 2. Mass spectrum of spent fuel from 130 uma to 160 u.
R. Brennetot et al. / Talanta 78 (2009) 676–681 Table 2 Yield obtained for Gd and Eu for an aliquot of 1 mL of synthetic solution after gravimetric separation by AG1X10 resin in HCl medium.
Aliquot 1 Aliquot 2 Aliquot 3
Eu yield
Gd yield
99 ± 2 99 ± 2 98 ± 2
99 ± 2 99 ± 2 98 ± 2
separation yield to be able to measure 147 Pm in spent fuels taking into account all the separation steps to correct the final measurement. It was decided to study the separation yield at each step of the analytical process used for 147 Pm measurement in order to find a marker of the separation to correct the final result of the separation yield. (1) Gravitational separation on resin: FP, U, Pu. (2) HPLC separation of actinides and lanthanides. (3) Measurement of the 147 Pm fraction free from 147 Sm by ICP-MS with a calibration curve at mass 147 with natural samarium. Ideally a double spike with the use of the isotope dilution technique would be optimal to overcome the measurement of the separation yields, but such a tracer is not available. It could be interesting to use another isotope of Pm as the 148 Pm but its lifetime is too much short to be used. Consequently, we decided to use a chemical analogue which would have the same chemical behaviour as Pm. The irradiated fuel is a complex solution which comprises many elements including lanthanides like shown in Fig. 2. Consequently, it is possible to use other lanthanides after ensuring that they have the same chemical behaviour as Pm with respect to the separation stages considered. Europium and gadolinium have been tested with respect to the separation stages (gravimetric separation + HPLC) in order to check their potential as tracer of the separation. We have chosen these two lanthanides for their proximity in mass and because they can be measured with a very small uncertainty, using the isotope dilution technique. This technique, well described in the literature [24] allows the determination of isotopic ratios and radionuclide concentrations with a very small uncertainty. 3.1. Gravimetric separation In order to analyze and to determine the content of 147 Pm in a spent nuclear fuel, it is necessary to separate the 147 Pm from the other elements such as fission products, U and Pu. This separation is carried out in two stages: the first separation is performed with an ion exchange resin in order to separate rapidly FP from the U and Pu matrix. Then, FP are separated by HPLC. In order to minimize the dose rate received by the experimenters, the yield of the first separation step for two selected lanthanides (Eu and Gd) was measured using a synthetic solution including only uranium, gadolinium and europium (natural) in proportions similar to those of spent nuclear fuels. This synthetic solution contains around 680 g/g U, 210 ng/g Eu and 390 ng/g Gd. Three tests were carried out. An aliquot of 1 mL of this synthetic solution is used, in order to make sure that a sufficient quantity of matter is able at the end of the first separation stages, to determine the yield by ICP-MS. Eu, and Gd concentrations are perfectly known as the synthetic solution is made by weighting. Concentrations in Eu and Gd were measured before and after gravimetric separation in order to determine the yield of separation and to prove that it could be considered as equivalent. The yields were measured for isotopes 151 and 153 of europium and 157 and 158 of gadolinium. The results from these three separations on resin are presented in Table 2. Error bars correspond to
679
reproducibility observed (4 calibrations ICP-MS with each time two test specimens of the fraction). At this stage, it can be considered that europium and gadolinium could be used like a tracer of the first separation stage since the separation yields are equivalent, very reproducible and close to 100%, suggesting therefore that no fractionation is induced by the gravimetric separation with the protocol chosen here. 3.2. HPLC A new synthetic solution containing 147 Pm at the concentration of 9 ng/g has been prepared. The 5 mL solution thus recovered at the end of the first separation previously described is put in a Teflon beaker and dried. The residue is then reconditioned in nitric medium with 5 mL of 0.2N Sub-boiling nitric acid compatible with HPLC. It is then injected on a HPLC chain confined in a glove box to separate Sm/Pm/Eu/Gd. The separation yield for Pm is obtained by calculation of the true concentration to date. The concentration in 147 Pm is measured by ICP-MS by using the 147 Sm as standard (it is supposed that Sm and Pm have the same response in ICP-MS). The results obtained on the three tests carried out in glove box are presented in Table 3. To ensure that the response of Pm and Sm could be expected equivalent in ICP-MS, the yield of separation for Pm is measured in the same solution by liquid scintillation counting, an independent measurement technique without any interference since the only beta emitter in the synthetic solution is 147 Pm. The results presented in Table 3 show that the same yields are obtained by liquid scintillation counting and by ICP-MS with samarium 147 used as a standard for external calibration for Pm measurement. This indicates at this stage, that europium and gadolinium could be used as tracer of the whole separation steps envisaged. 3.3. Analysis of spent fuels Previous developments with synthetic solutions show that the yield at the end of the separation stages with selected conditions are equivalent for selected lanthanides Pm, Eu, Gd. Thus 153 Eu and 157 Gd could be used as tracers of the separation yield. In addition it has been shown that 147 Sm could be used to carry out the range of calibration to measure 147 Pm. Two spent fuel solutions were analyzed. Separations were carried out on the MOx and UOx fuels. Two aliquots of 250 l for each fuel were taken for separation steps. Chromatogram of one of the aliquot of UOx sample is presented in Fig. 3. As shown in Fig. 2, there is lots of peaks in the lanthanide zone for a spent fuel what results in to create interferences which were not taken into account up to now. Finally 153 Eu was chosen as the tracer of the separation, because of the presence of some oxides (BaO, CeO, PrO) that induce an over estimation of the absolute content of gadolinium in the fuel without separation step. For this purpose, it has been experimentally verified that caesium do not produce a sufficient oxide quantity to interfere at mass 149, Table 3 Yield obtained for Gd, Eu and Pm for an aliquot of 1 mL of synthetic solution after gravimetric separation and HPLC. The first column for Pm is the yield determined by ICPMS with Sm external calibration, the second one is the determination by liquid scintillation counting.
Aliquot 4 Aliquot 5 Aliquot 6
Eu yield
Gd yield
Pm yield
81 ± 3 71 ± 3 81 ± 3
80 ± 3 73 ± 3 83 ± 3
81 ± 3 70 ± 3 77 ± 3
79 ± 3 70 ± 3 78 ± 3
680
R. Brennetot et al. / Talanta 78 (2009) 676–681
returned by isotopic dilution with a variation observed lower than 1% for UOx sample and less than 3% for MOx sample. The 147 Pm concentration obtained were then compared with prediction calculation according to DARWIN 2.2 code with JEFF 2.2 library. Less than a 10% variation is obtained between the measurement and the result obtained starting from the computer code. However this variation is covered by the uncertainty of measurement which validates this protocol. 4. Conclusion
Fig. 3. Chromatogram obtained with radiometric detection for separation of the FP fraction obtained at the first gravimetric separation of UOx GGU1 fuel.
150, 151, 153 when barium, praseodymium and cerium produces a large amount of oxides that interferes all the isotopes of gadolinium. The results are presented in Table 4. Uncertainties were given on several levels in order to try to take into account all stages leading to the final result. The analytical uncertainty given in ICP-MS takes into account all the analytical stages, of the preparation of the sample received at the laboratory with its analysis: preparation of the standards by weighing, uncertainty on the concentrations of the standards (calculated starting from the certificates of analysis of the standard solutions), uncertainty on the calibration, stability of the apparatus, estimate of the accuracy of the analyzes starting from standards whose concentration is known, variation of measurements on two distinct test specimens, derives possible from the apparatus over the duration of measurements. Uncertainties on the concentrations determined in ICP-MS were given starting from LANI, homemade software which integrates the sources of uncertainties such as dilutions, uncertainty on the standards and the hyperboles of confidence of the calibration line. The values of concentrations in 153 Eu obtained before and after separations are compiled in order to obtain the output of the stages of separation. Resulting uncertainty is given with the method described by Kragten [25]. Combined uncertainty is calculated starting from the law of propagation of uncertainties like defined in EURACHEM/CITAC guide [26]. The result obtained on the absolute concentration of 153 Eu was first of all compared with the concentration returned by isotopic dilution (ID) in June 2006 (taken as reference value). As can be seen in Table 4, europium concentration is coherent with the result Table 4 Results obtained for 153 Eu and 147 Pm concentration in spent fuels. Separation yields applied for 147 Pm concentration are calculated from 153 Eu concentration measured by Q-ICP-MS before and after separation steps. 153
Eu (ng/g sol)
153
Eu (ng/g sol) reference value
147
Pm (ng/g sol)
UOx Before sep steps: 161.7 After sep steps: 128.3
163.3 ± 3.2
27.5 ± 2.7
MOx Before sep steps: 266.4 After sep steps: 217.5
273.2 ± 5.9
40.14 ± 4.06
With an aim of determining the concentration in 147 Pm in UOx and MOx spent fuel solutions, a durable method was developed in order to validate the separation stages necessary to this measurement (exchange resin and HPLC). This new measurement was optimized to complete a method previously used for lanthanides determination by isotope dilution technique. These tests made it possible to show that lanthanides Eu, Gd and Pm had a similar behaviour with respect to the selected resin (AG1X10) in hydrochloric medium in this type of matrix. We have shown that Eu and Gd had the same behaviour during all the separations steps, making it possible their use as a tracer of the separation for Pm measurement. Checking of the yield at the conclusion of the stages of separation by liquid scintillation made it possible to validate on the one hand the yields determined by ICP-MS, but also to validate the use of the natural 147 Sm to measure the 147 Pm after separation. The yield thus determined for the 153 Eu could be applied to the measurement of the 147 Pm in fuel. These results also were validated by ICP-MS with a difference lower than 2% compared to the results resulting from the determinations by isotopic dilution in May and October 2006 to the measurements of absolute concentrations of the isotope 153 Eu used for the follow-up and the determination of the yield on the totality of the experiment. The measurements obtained in ICP-MS for the 147 Pm and measurements of output of 153 Eu made it possible to calculate the concentrations in 147 Pm in fuel. These results are completely coherent with the values returned by calculation code with a difference less than 10% and totally covered by uncertainty measurement. Acknowledgments The authors wish to thank Cecile Riffard (CEA Cadarache, DER/SPRC/LECy) for its support for calculation codes. Authors wish to thank also Carole frechou (DEN/DANS/DPC/SECR/LANIE) for her advices and Pascal Fichet (DEN/DANS/DPC/SECR/LANIE) for his uncertainty determination program. References [1] B. Roque, N. Thiollay, P. Marimbeau, A. Barreau, A. Tsilanizara, C. Garzenne, F. Marcel, H. Toubon, C. Garats, Park, MG, (Ed.), Proceedings PHYSOR-2002, Seoul, Korea, 7–10 October 2002. [2] S.M. Jerome, Sci. Total Environ. 70 (1988) 275. [3] M. Yoshida, S.H. Sumiya, H. Watanabe, K. Tobita, J. Radioanal. Nucl. Chem. 197 (1995) 219. [4] G. Seeber, P. brunner, M.R. Buchmeiser, G.K. Bonn, J. Chromatogr. A 848 (1999) 193. [5] S. Eichuk, C.A. Lucy, K.I. Burns, Anal. Chem. 64 (1992) 2339. [6] L. Adriaensen, M. Gysemans, C. Hurtgen, D. Boulanger, ENC 2007 Conference, Brussels, September 16–20, 2007. [7] F. Chartier, M. Aubert, M. Salmon, M. Tabarant, B.H. Tran, J. Anal. Atom. Spectrom. 14 (1999) 1661. [8] F. Chartier, M. Aubert, M. Pilier, Fresenius J. Anal. Chem. 364 (1999) 320. [9] H. Isnard, R. Brennetot, C. Caussignac, N. Caussignac, F. Chartier, Int. J. Mass Spectrom. 246 (2005) 66. [10] I. Gunther Leopold, N. Kivel, J. Kobler, B. Wernli, Anal. Bioanal. Chem. (2008) 503. [11] H. Isnard, M. Aubert, P. Blanchet, R. Brennetot, F. Chartier, V. Geertsen, F. Mannuguerra, Spectrochim. Acta 61B (2006) 150.
R. Brennetot et al. / Talanta 78 (2009) 676–681 [12] J.S. Becker, W. Kerl, H.J. Dietze, Anal. Chim. Acta 387 (1999) 145. [13] D. Solatie, P. Carbol, M. Betti, F. Bocci, T. Hiernaut, V.V. Rondinella, J. Cobos, Fresenius J. Anal. Chem. 368 (2000) 88. [14] C.S. Kim, C.K. Kim, P. Martin, U. Sansone, J. Anal. Atom. Spectrom. 22 (2007) 827. [15] J.I. Garcia Alonso, F. Sena, P. Arbore, M. Betti, L. Koch, J. Anal. Atom. Spectrom. 10 (1995) 381. [16] J. Comte, P. Bienvenu, E. Brochard, J.M. Fernandez, G. Andreoletti, J. Anal. Atom. Spectrom. 18 (2003) 702. [17] J.I. Garcia Alonso, J.F. Babelot, J.P. Glatz, O. Cromboom, L. Koch, Radiochim. Acta 62 (1993) 71. [18] J.S. Becker, R.S. Soman, K.L. Sutton, J.A. Caruso, H.J. Dietze, J. Anal. Atom. Spectrom. 14 (1999) 933. [19] S.F. Boulyga, C. Testa, D. Desideri, J.S. Becker, J. Anal. Atom. Spectrom. 10 (1995) 381.
681
[20] N. Baglan, C. Bouvier Capely, C. Cossonnet, Radiochim. Acta 90 (2002) 267. [21] P. Roos, H. Xiaolin, Anal. Chim. Acta 608 (2008) 105. [22] M. Marhol, Ion Exchangers in Analytical Chemistry. Their Properties and Use in Inorganic Chemistry, vol. XIV, Elsevier Scientific Publishing Company, 1982, 585 p. [23] G. Favre, R. Brennetot, F. Chartier, P. Vitorge, Int. J. Mass Spectrom. 265 (2007) 15. [24] K.G. Heumann, Int. J. Mass Spectrom. 118/119 (1995) 575. [25] J. Kragten, Analyst 119 (1994) 2161. [26] EURACHEM Secretariat, EURACHEM/CITAC Guide Quantifying Uncertainty in Analytical Measurement, second ed., EURACHEM Secretariat, BAM, Berlin, 2000.