Determination of 107Pd in Pd Recovered by Laser ... - ACS Publications

Nov 25, 2016 - by pulsed laser irradiation at 355 nm provides short-time and one-step recovery of Pd. The proposed method was verified by applying it ...
2 downloads 8 Views 882KB Size
Download PDF
Article pubs.acs.org/ac

Determination of 107Pd in Pd Recovered by Laser-Induced Photoreduction with Inductively Coupled Plasma Mass Spectrometry Shiho Asai,*,† Takumi Yomogida,† Morihisa Saeki,‡ Hironori Ohba,‡ Yukiko Hanzawa,† Takuma Horita,† and Yoshihiro Kitatsuji† †

Nuclear Science and Engineering Center, Japan Atomic Energy Agency, Ibaraki 319-1195, Japan Quantum Beam Science Research Directorate, National Institutes for Quantum and Radiological Science and Technology, Ibaraki 319-1106, Japan

‡

S Supporting Information *

ABSTRACT: Safety evaluation of a radioactive waste repository requires credible activity estimates confirmed by actual measurements. A long-lived radionuclide, 107Pd, which can be found in radioactive wastes, is one of the difficult-to-measure nuclides and results in a deficit in experimentally determined contents. In this study, a precipitation-based separation method has been developed for the determination of 107Pd with inductively coupled plasma mass spectrometry. The photoreduction induced by pulsed laser irradiation at 355 nm provides short-time and one-step recovery of Pd. The proposed method was verified by applying it to a spent nuclear fuel sample. To recover Pd efficiently, a natural Pd standard was employed as the Pd carrier. Taking advantage of the absence of 102Pd in spent nuclear fuel, 102Pd in the Pd carrier was utilized as the internal standard. The chemical yield of Pd was about 90% with virtually no impurities, allowing accurate quantification of 107Pd. The amount of 107Pd in the Pd precipitate was 17.3 ± 0.7 ng, equivalent to 239 ± 9 ng per mg of 238U in the sample.

P

0.5 M HCl and readily eluted with 8 M HNO3, giving a chemical yield ranging from 70%−89%. The method was applied to radioactive wastes, such as evaporator concentrate and radioactive sludge sampled from damaged nuclear facilities that were severely contaminated with fission products originating from partially melted fuel. However, none of the results exceeded the minimum detectable activity of LSC. Other than the low specific activity of 107Pd, the instability of trace Pd ions in an aqueous solution may have some connection to the difficulties in detecting 107Pd in radioactive wastes. Several reports showed that Pd is likely to form a precipitate that can be incorporated into insoluble residue in an HNO3 matrix.1,4,5 This makes it difficult to track the migration path of dissolved species in waste processing. Mass spectrometry is advantageous over radiometry in measuring long-lived radionuclides that have low specific activity, achieving low detection limits. A sequential separation scheme for the determination of 93Zr, 135Cs, and 107Pd using TBP-extraction chromatography, and anion-exchange chromatography combined with electro-thermal vaporization inductively coupled plasma mass spectrometry (ETV-ICP-MS), was applied to a zircaloy hull solution.6 The chemical yield and

alladium isotopes are generated during nuclear reactions along with many other fission products. Both radio and stable isotopes can be found in spent nuclear fuel and high-level radioactive waste (HLW).1,2 Among these isotopes, 107Pd is the only long-lived nuclide (half-life: 6.5 × 106 y). Thus, the determination of the 107Pd content in radioactive waste is considered crucial for safety evaluation in managing such wastes. To establish a safe and cost-effective design for the repository, reliable and precise estimation ensured by the actual concentrations of 107Pd in each waste package is essential, rather than overestimations based on the limits of detection. However, to the best of our knowledge, experimentally determined concentrations of 107Pd in wastes have not been reported. The difficulties in the determination of 107Pd are mainly associated with its radiochemical properties. Palladium-107 is a pure β-emitter with decay energy of 33 keV. To identify such low-energy β-rays, complete removal of the coexisting radionuclides is a prerequisite to avoid spectral interference. In addition, because of its low specific activity, long-time measurement is necessary to achieve precise quantification. Dulanská et al. proposed a separation method based on extraction chromatography using a commercially available resin (Ni Resin)3 for liquid scintillation counting (LSC). Dimethylglyoxime (DMG) is impregnated inside the pores of an inert porous polymer. The Pd ions were extracted as Pd(DMG)2 in © XXXX American Chemical Society

Received: August 22, 2016 Accepted: November 25, 2016 Published: November 25, 2016 A

DOI: 10.1021/acs.analchem.6b03286 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry detection limit of 107Pd were 60% and 0.4 μg/g-Zr, respectively. No quantified data were shown in the report.6 As indicated by the previous measurements,3,4,6 spent nuclear fuel that is predicted to contain 107Pd at significant levels can be an ideal sample for 107Pd quantification. Besides the methods for radio palladium measurement, several separation techniques based on coprecipitation7−9 and column chromatography10,11 have been proposed for stable Pd determination. Te coprecipitation was applied to the preconcentration and purification of Pd, along with other platinum group metals (PGMs), in geological7 and environmental8,9 samples. Te coprecipitation generally comprises at least two procedures: heating of a sample solution with added Te and the addition of reducing agents to the preheated solution. Sn(II) chloride7,8 is employed as the reducing agent. Elements that cause spectral interference in the mass region of Pd isotopes, such as Cu, Rb, Sr, Y, Zr, Mo, and Cd are efficiently removed. An automated preconcentration−purification technique using online columns loaded with complexation agents, for example, N,N-diethyl-N′-benzoylthiourea10 and 1,5-bis (2pyridyl)-3-sulphophenyl methylene thiocarbonohydrazide,11 was successfully applied to the determination of Pd in environmental samples. This enables the possibility of inadvertent contamination by foreign substances to be reduced. Although the methods described above offer high chemical yields and decontamination factors for all the potentially interfering elements, they are inapplicable to the determination of 107Pd, which is inevitably associated with the treatment of highly radioactive sample, or spent nuclear fuel. To achieve safe and accurate quantification of 107Pd, a much simpler separation method is required to reduce risks associated with the operation-related radiation exposure and radioactive contamination of instruments and laboratory equipment. This study focuses on the development of a specialized separation method for highly radioactive samples using photoreduction induced by laser irradiation, which is a completely different approach12−15 from those described in previous studies. This laser-induced precipitation technique enables a short-time and one-step recovery of Pd in closed vessel at room temperature. Preceding studies showed the photoreduction conditions of Pd ions using laser irradiation in an HNO3 matrix.12 The purpose of the study was to evaluate the applicability of photoreduction induced by laser irradiation to PGM recovery. Thus, Pd solutions prepared within a high concentration range (approximately mmol/L level) were used for the experiments. The irradiation source employed in the experiments was a nanosecond pulsed Nd:YAG laser operating with a wavelength and pulse length of 355 nm and 10 ns, respectively. The laser energy was set at 80 mJ/pulse. During the laser irradiation, Pd(II) ions are reduced to change into the metal state Pd(0) and the resulting metal cores grow large enough to form a precipitate.12−14 The UV−vis spectra of the Pd chloride solutions before and after irradiation suggested that Pd ions in 1 M HNO3 were efficiently reduced in the presence of oxalic acid. The resulting Pd precipitate was readily collected on a quantitative filter paper. The formation of pure Pd metal after irradiation was confirmed by comparing the XRD spectra of the pure metal Pd standard with that of the collected Pd precipitate. However, there have been no reports that describe the applicability of the laser-induced precipitation technique to

the purification and preconcentration of Pd for its quantification. Our previous studies15 additionally demonstrated that Pd ions were selectively extracted from a simulated HLW solution with a similar laser setting to that in the aforementioned work.12 The simulated HLW solution was prepared in accordance with the typical chemical composition of spent nuclear fuel. Before irradiation, the solution was diluted with an HNO3−ethanol mixed solution to adjust the concentration of Pd to around 2.5 μg/mL. Here, ethanol acts as the reducing agent.13 The concentrations of HNO3 and ethanol in the final solution for laser irradiation were set at 0.5 mol/L and 40 v/v%, respectively. Approximately 60% of the Pd was recovered after 20 min irradiation with almost no impurities.15 This suggests that highly selective separation of Pd is achievable with the laser-induced precipitation technique. However, the quantitative photoreduction of Pd is typically limited to a solution containing Pd at a concentration greater than 2.5 μg/mL, depending on the irradiation settings.15 To prepare a Pd solution at the μg/mL level, a spent nuclear fuel sample solution with high radioactivity is required. The predicted content of Pd in spent nuclear fuel provided by a burnup calculation code, ORIGEN2, is approximately 2.5 μg per mg of U,2 which has a total activity of more than 10 MBq. Handling such high-activity samples in a special facility equipped with radiation shielding imposes many restrictions on operations. In this study, to eliminate such limitations on the Pd concentration, a natural Pd standard was employed as the Pd carrier. The natural Pd carrier is identifiable in the spent nuclear fuel sample because Pd found in spent nuclear fuel has a completely different isotopic composition than that of natural Pd. By taking advantage of the absence of 102Pd in spent nuclear fuel,16 102Pd in the Pd carrier is usable as the internal standard for ICP-MS. With such an internal standard, the measurements of Pd can be free from the losses of analyte Pd during the separation process, signal drifts, and matrix effects. The Pd concentration is adjustable to a feasible level for photoreduction (around 2.5 μg/mL) by controlling the amount of the Pd carrier added to the sample as well as the amount of the sample itself. The total activity of the sample can be suppressed by increasing the contribution of the Pd carrier against the sample. However, an excessive amount of the carrier may cause spectral interference at m/z 107 with 106Pd1H+. Thus, the weight ratio of the Pd carrier to the Pd initially existing in the sample needs to be optimized before applying the laser irradiation technique to the 107Pd determination of a real sample. Our study is the first attempt to measure the concentration of 107Pd in spent nuclear fuel. The proposed method was verified using a sample with a traceable irradiation history. The chemical yields of the major components were measured to confirm that the high selectivity for Pd can be reproduced in the spent nuclear fuel sample, which is highly radioactive and predominantly comprises U (≈94%) along with many other fission products and actinides. The interference at m/z 107, attributable to spent nuclear fuel, was also evaluated.

■

EXPERIMENTAL SECTION Reagents. Natural Pd standard solutions in 2%−3% HNO3 and 1 M HCl were supplied by Merck KGaA and Wako Pure Chemical Industries Ltd., respectively. A rare earth element-mix standard solution (Ce, Dy, Er, Eu, Gd, Ho, La, Lu, Nd, Pr, Sc, Sm, Tb, Tm, U, Y, and Yb at 10 mg/L in 2v/v% HNO3) and B

DOI: 10.1021/acs.analchem.6b03286 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

and -3 for solutions before irradiation; RP-1, -2, and -3 for the Pd precipitates). The measured concentrations of the components in each prepared solution, including the 105Pd carrier, are summarized in Table 1, along with the sample matrix conditions. The sample volumes were calculated with the weights of the sample solution and the densities.

multielement standard solution (XSTC-331) were obtained from High-Purity Standards and SPEX CertiPrep, respectively. Other single-element standard solutions with concentrations of 1000 mg/L were purchased from Wako Pure Chemical Industries Ltd. A Pd-105 enriched metal standard in powder form (105Pd: 96.58%) was obtained from Oak Ridge National Laboratory (ORNL). HNO3 and HCl used for chemical separation and preparation of final solutions for ICP-MS measurements were of ultrapure grade (TAMAPURE AA-10 for HNO3 and AA-100 for HCl) and were supplied by TAMA Chemicals. Ultrapure water with a resistivity of 18.2 MΩ·cm prepared with a Milli-Q system was used throughout the experiments. Analytical grade ethanol was used without further purification. Model Solution. A model solution that simulates spent nuclear fuel was prepared by Wako Pure Chemical Industries Ltd. and used for confirming the validity of the proposed photoreduction conditions. The chemical composition of the model solution was decided in accordance with spent nuclear fuel with a typical irradiation history.2 For easier operation, elements that have no stable nuclides, such as Tc, U, Np, Pu, and Am, were excluded. Prescribed amounts of the components (Ba, Ce, Cs, La, Mo, Nd, Pd, Pr, Rb, Rh, Ru, Sm, Sr, and Zr) were dissolved in dilute HNO3. The concentrations of HNO3 and Pd were set at 3 mol/L and 480 μg/mL, respectively, which allows the stability of the prepared model solution to be maintained. The concentration of Pd in the model solution corresponds to that in spent nuclear fuel solution at approximately 300 mg of U per mL. The concentrations of each metal in the prepared model solution were verified by ICP-MS before use. To identify the carrier Pd from the Pd in the model solution, the 105Pd enriched standard solution was employed as a carrier. About 5 mg of the 105Pd standard powder was dissolved in 0.1 mL of aqua regia at room temperature and diluted with 5 mL of 7 M HNO3. The concentration of 105Pd in the 105Pd-enriched standard solution was determined on the basis of reverse isotope dilution mass spectrometry using a natural Pd standard solution as a spike.16 The resulting 105Pd standard solution was diluted to adjust its concentration to 190 μg/mL with 7 M HNO3. The model spent nuclear fuel solution was further diluted to adjust the concentration of Pd to 660 ng/mL with 7 M HNO3. Immediately before laser irradiation, the diluted model solution (0.16 mL, corresponding to roughly 0.07 mg of U in spent nuclear fuel) and 105Pd carrier solution (0.05 mL) were mixed in a screw-top quartz cell (12 mm × 12 mm × 40 mm, thickness: 1 mm) with a plastic cap. Subsequently, 1.8 mL of pure water and 1.3 mL of ethanol were added to the cell. HNO3 in this case was employed as a solvent instead of HCl and HClO, which were also used in the previous studies.12,13 In facilities for highly radioactive samples, the use of corrosive acids, such as HCl and HClO, is considered to be inappropriate. Additionally, our early studies showed that Rh and Ru ions in HCl were reduced to their metal state together with Pd ion.13 This erodes the selectivity for Pd. For the reducing agent, we judged that ethanol was the most suitable in view of managing the radioactive liquid waste left after reaction. Ethanol can be removed from the postirradiated solution by evaporation, thereby making it easier to stabilize the waste solution as compared to other agents with higher boiling points. To evaluate the reproducibility of this method, the same procedure was repeated three times (sample names: RS-1, -2,

Table 1. Measured Concentrations (ng/g) of Components in Model Solution along with Sample Matrix Conditions elements

RS-1

RS-2

RS-3

Ba Ce Cs La Mo Nd Pr Rb Rh Ru Sm Sr Zr natural Pd (sample) 105 Pd (carrier) Pd-total sample matrix volume (mL) HNO3 (mol/L) ethanol (v/v%)

39.3 60.5 68.7 31.0 83.2 101.8 28.9 9.2 10.8 54.5 21.8 19.9 88.5 32 2737 2769

36.5 55.2 63.1 28.7 77.2 93.3 26.4 8.5 10.1 50.0 20.6 18.5 81.0 30 3030 3059

35.3 55.8 62.1 28.4 76.5 93.1 26.2 8.2 9.3 50.3 20.3 18.1 81.3 29 3052 3081

3.41 0.45 0.39

3.43 0.47 0.39

3.44 0.46 0.39

Spent Nuclear Fuel Sample. A spent nuclear fuel solution was prepared by dissolving a single fuel pellet irradiated in a Japanese PWR with a burnup of 44.9 GWd/t. The cooling period is 10 257 days. The weight, excluding the zircaloy cladding, was approximately 5 g. The irradiation history and dissolution procedures were detailed in our previous publications.17−19 A burnup calculation code, ORIGEN2, provides the abundance ratio of Pd to U (238U). Accordingly, the Pd concentration in every sample solution prepared by diluting the original spent nuclear fuel solution can be estimated using ORIGEN2 with a nuclear data library, JENDL 4.0, and the measured 238U concentrations. Although the reliability of the abundance ratio, Pd/238U, given by ORIGEN2 has not been confirmed, the predicted ratio is still helpful for adjusting the concentration of Pd to a feasible level for photoreduction. The insoluble residue formed in the resulting solution was removed by filtration and the supernatant was used in this study. Part of the supernatant was diluted with 7 M HNO3 to adjust the concentration of Pd to be suitable for laser irradiation (0.4 mg of U per mL). Moreover, another solution was prepared by further diluting the supernatant with 1 M HCl for ICP-MS. The concentration of U in the solution was set at approximately 700 ng of U per mL. The measured and predicted concentrations of the major nuclide found in the spent nuclear fuel diluted solution are listed in Table 2. The predicted concentrations were provided through ORIGEN2. The measured concentrations of Ba, Cs, Rb, Sr, and U, which stably exist in the nitric acid matrix, generally agreed with the predicted values. In contrast, parts of Mo, Pd, Ru, and Rh appeared to be lost by incorporation in the insoluble residue C

DOI: 10.1021/acs.analchem.6b03286 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry Table 2. Measured and Theoretically Calculated Concentrations of Diluted Spent Nuclear Fuel Solution nuclide 85

Rb Sr 91 Zr 95 Mo 101 Ru 103 Rh 133 Cs 107 Pd* 139 La 140 Ce 141 Pr 146 Nd 147 Sm 238 U 88

diluted spent nuclear solution (ng/g)

predicted by ORIGEN2 (ng/g)

± ± ± ± ± ± ± ± ± ± ± ± ± ±

0.12 0.33 0.58 0.78 0.85 0.47 1.13 0.30 1.29 1.33 1.19 0.79 0.23 653

0.11 0.33 0.58 0.71 0.41 0.23 1.12 0.16 1.31 1.32 1.21 0.78 0.22 650

0.01 0.01 0.05 0.06 0.03 0.02 0.01 0.02 0.05 0.05 0.03 0.04 0.01 11

Figure 2. Pd separation procedure based on laser-induced precipitation technique.

for ICP-MS. The chemical yield of each component in the model solution was calculated using the following equation: chemical yield [%] = AP /(AS + A C) × 100

*

(1)

where AP and AS are the amounts of an element of interest found in the Pd precipitate and in the model solution added to the cell, respectively. Only for Pd, the added amount of Pd in the 105Pd carrier, AC, was considered. Both AP and AS were obtained from the measured concentrations of each component. For the spent nuclear fuel sample, the chemical yield of Pd was calculated from the weight of 102Pd in the Pd precipitate for the numerator and the weight of 102Pd in the natural Pd carrier of the cell for the denominator. Regarding the other components, each chemical yield of an element found in the Pd precipitates was calculated using eq 1. For Tc and actinides (Np, Pu, Am, and Cm), the net count rates obtained by ICPMS in the Pd precipitates and the sample solutions were used instead of the amounts of the element of interest. ICP-MS. A quadrupole ICP-MS Agilent 7700x was used for all the measurements. The operation conditions were basically set according to our previous study.19 The final solutions were prepared with 1 M HCl. The sensitivity at m/z 105 was 290 000 cps/ng g−1 (RSD: 1%−2%). The formation rates of oxide ions (140Ce16O+/140Ce) and doubly charged ions (140Ce2+/140Ce) were controlled daily so as not to exceed 1.5% and 2.0, respectively, using a tuning solution (1 μg of Li, Y, and Ce mixed solution). Under the above-described conditions, the average formation rate for 91Zr16O+, which would be a main interference in 107Pd measurement, was 0.5%. The hydride formation rate of Pd at m/z 107 (106Pd1H+/106Pd) was 10 MBq). A natural Pd carrier enables quantitative Pd precipitation, even if a much smaller amount of spent nuclear fuel sample is applied, compensating for the insufficient amount of Pd in the original sample. However, 106 Pd (natural abundance: 0.2733) in the Pd carrier may cause spectral interference in 107Pd determination, forming 106Pd1H+. Assuming a 100% natural Pd solution at 2.5 μg/g (0.68 μg-106Pd/g), the estimated additional cps (106Pd1H+) at m/z 107 is around 2000 when considering the measured hydride formation rate (106Pd1H+/106Pd+ < 0.001%), and the sensitivity for Pd (290 000 cps/ng g−1). As the RSD of the measurements at m/z 107 was at least 1%, a 107Pd signal rate of 200 000 cps, equivalent to 0.6 ng-107Pd/mL, is the practically minimum applicable level to cancel the interference caused by 106Pd1H+. The concentration of the total Pd isotopes originating from spent nuclear fuel at such a minimum applicable level is theoretically estimated to be 4 ng/mL, corresponding to approximately 1/600 of the feasible level for photoreduction (2.5 μg/mL). In this study, considering the loss of Pd during the sample preparation procedure, the provisional Pd concentration in the spent nuclear fuel sample was set at 1/ 100 (25 ng/mL) of that in the Pd carrier. It can be noted that the Pd concentrations discussed above are assumed figures. The calculated count rates are intended for use only in the evaluation of the relative relationship between counts at m/z 106 and 107. Solutions diluted by a factor of 10 or higher were actually measured to avoid severe memory effects. Evaluation of Separation Performances Using Model Solution. The separation performances of the laser-induced precipitation technique were evaluated using a nonradioactive model solution. The 105Pd-enriched standard solution, rather than the natural Pd standard solution was mixed with the model sample solution as a Pd carrier. The Pd concentration in the model solution was set at approximately 1/100 of that in the Pd carrier by adjusting the added volumes of the model sample solution and the 105Pd standard solution to 0.16 mL (660 ng/mL) and 0.05 mL (190 μg/mL), respectively. Pd precipitation was obtained by following the procedure shown in Figure 2. The chemical yields for each component in the Pd precipitates are listed in Table 3. Constant yields of 80% for Pd were obtained, suggesting that the proposed photoreduction conditions (laser irradiation parameters, composition of the matrix solution, and concentration of Pd) are preferable for stable formation of the Pd precipitate. For the other components, the amounts incorporated into the Pd precipitates were negligibly low, thus demonstrating a high selectivity for Pd. The procedure blanks summarized in Table 4 indicate that none of the potentially interfering isotopes were found at significant levels. In particular, the negligibly low contamination

element

RP-1

RP-2

RP-3

Pd Ba Ce Cs La Mo Nd Pr Rb Rh Ru Sm Sr Zr

80.4 0.08 0.03