Article pubs.acs.org/ac
Determination of Atto- to Femtogram Levels of Americium and Curium Isotopes in Large-Volume Urine Samples by Compact Accelerator Mass Spectrometry Xiongxin Dai,*,† Marcus Christl,‡ Sheila Kramer-Tremblay,† and Hans-Arno Synal‡ †
Canadian Nuclear Laboratories, Chalk River, Ontario K0J 1J0, Canada Laboratory of Ion Beam Physics, ETH Zurich, 8093 Zurich, Switzerland
‡
S Supporting Information *
ABSTRACT: Ultralow level analysis of actinides in urine samples may be required for dose assessment in the event of internal exposures to these radionuclides at nuclear facilities and nuclear power plants. A new bioassay method for analysis of subfemtogram levels of Am and Cm in large-volume urine samples was developed. Americium and curium were co-precipitated with hydrous titanium oxide from the urine matrix and purified by column chromatography separation. After target preparation using mixed titanium/iron oxides, the final sample was measured by compact accelerator mass spectrometry. Urine samples spiked with known quantities of Am and Cm isotopes in the range of attogram to femtogram levels were measured for method evaluation. The results are in good agreement with the expected values, demonstrating the feasibility of compact accelerator mass spectrometry (AMS) for the determination of minor actinides at the levels of attogram/liter in urine samples to meet stringent sensitivity requirements for internal dosimetry assessment.
I
239
Pu/240Pu, 243Cm/244Cm, and 245Cm/246Cm). In these cases, development of ultrasensitive mass spectrometric methods is critical to shorten the sample analysis time and to enhance the analytical sensitivity. Accelerator mass spectrometry is an extremely sensitive and robust technique for the analysis of intermediate- and longlived radionuclides.8−13 As a result of its high rejection of molecular isobaric interferences and low susceptibility to matrix effects, the AMS technique allows for simplification of the sample preparation chemistry with a good potential for high sample analysis throughput and reduced cost for ultralow level radioassays. This is of particular interest for the ultratrace determination of higher actinides (especially Am and Cm nuclides) in bioassay samples, where other mass spectrometric methods (such as inductively coupled plasma mass spectrometry (ICP-MS) and thermal ionization mass spectrometry (TIMS)) may not be adequate to meet analytical sensitivity or throughput requirements. Although AMS has been demonstrated for the precise and accurate determination of Pu at fg-levels in urine bioassay samples,14−16 the use of this technique for routine measurements has been limited by lack of availability and high operational costs of the complicated AMS system. In our recent study, the applicability of a compact AMS system at the Swiss Federal Institute of Technology (ETH) has been demonstrated
nternal exposure of nuclear workers to actinides could occur during decommissioning activities of nuclear facilities, refurbishment of aging reactors, handling of spent nuclear fuel, and reprocessing of radioactive wastes. Highly radiotoxic alpha emitting actinides, such as 241Am, 244Cm, and 239/240Pu, are likely to be the major contributors of the internal dose, as these nuclides are present in significant amounts in irradiated fuels, particularly high burn-up spent fuels.1,2 Therefore, ultralow level analysis of higher actinides in bioassay samples is often required for dose assessment in the event of such accidental exposures.3,4 Due to its easy collection, urinalysis is the most commonly used in vitro bioassay method for internal contamination monitoring of actinide exposure. However, this often requires extremely sensitive analytical techniques to detect actinides at low femtogram levels or less in urine samples (see Table 1), since the detection of all exposures that may exceed a committed effective dose (CED) of 1 mSv per year shall be ensured through bioassay monitoring as recommended by the International Commission on Radiological Protection (ICRP) and the nuclear safety authorities in many countries, including the Canadian Nuclear Safety Commission (CNSC).5 Although high resolution alpha spectrometry has been the preferred method for measuring alpha emitting radionuclides with half-lives shorter than 1000 years (e.g., 241Am and 242,243,244 Cm), a very long counting time (often a few weeks) is needed to achieve a minimum detectable activity (MDA) of 15 min and transferred to a centrifuge bottle. After centrifugation, the supernatant solution was decanted. The precipitate was then dissolved and transferred to a Teflon beaker with ∼10 mL of concentrated HNO3. The sample was heated to boiling with 0.5 mL of 30% H2O2 to decompose the residual organics. The actinides were co-precipitated with HTiO again by neutralization with NH4OH, and the supernatant solution was decanted after centrifugation. The HTiO precipitate was rinsed with water and centrifuged to remove residual salt. In the subsequent step, the precipitate was dissolved in an equal volume of concentrated HNO3 to make up to a final acidity of 8 M nitric acid and 0.5 mL of H2O2 was added prior to chromatographic column separation. The chromatographic columns used for actinide separation were a 2 mL Eichrom TEVA cartridge (for the removal of Pu and Np) stacked on top of a 2 mL DGA cartridge (for the extraction of Am and Cm). After preconditioning with 10 mL of water followed by 10 mL of 8 M HNO3, the sample was passed through the columns at a rate of ∼1 mL·min−1. The columns were then rinsed with ∼20 mL of 8 M HNO3 and separated for elution. The Am/Cm/Cf extracted onto the DGA resin was eluted using 13 mL of 0.05 M HCl for the preparation of AMS target. AMS Target Preparation. The AMS target was prepared by an optimized method using mixed titanium and iron hydroxide co-precipitation. To do this, 0.4 mg of Ti and 0.1 mg of Fe from standard solutions was added to the eluate, and the sample was neutralized with concentrated NH4OH to pH > 9 to co-precipitate the Am/Cm/Cf. The supernatant solution was decanted after centrifugation and the precipitate was transferred with methanol to a 1 mL microcentrifuge tube. It was further rinsed with methanol and centrifuged prior to drying in a heating block. The dried sample was finally mixed with 4−5 mg of niobium powder and pressed into a Ti target holder for AMS measurement. AMS Analysis. The AMS measurements were performed using the low energy (0.6 MV) AMS system TANDY at ETH Zurich. The details about the AMS setup for actinide measurements have been described elsewhere.19 To summarize, negatively charged actinide oxide ions (AnO−) ions were extracted from the Cs-sputter ion source and injected into the accelerator running at a terminal voltage of about 320 kV. At
Figure 1. Flow diagram of urine bioassay procedure for americium/ curium.
of 243Am, 243Cm, 248Cm (252Cf) tracer was added to 1.6 L of acidified urine or 400 mL of reagent blank in a glass beaker for procedural efficiency correction (see Table 2). The sample was well mixed at
