Journal of Radioanalytical and Nuclear Chemistry, Vol. 243, No. 2 (2000) 397 401 On the use of ICP-MS for measuring plutonium in urine N. Baglan, 1 C. Cossonnet, 1 P. Pitet, 1 D. Cavadore, 2 L. Exmelin, 3 P. Berard 1 1 Institut de Protection et de Sureté Nucléaire, Département de Protection de la Santé de l Homme et de Dosimétrie, Service de Dosimétrie, IPSN, BP n 6, F 92265 Fontenay-aux-Roses Cedex, 2 COGEMA, LAM Marcoule, BP170, F 30206 Bagnols sur Ccze Cedex, 3 COGEMA, LAM La Hague, F 50444 Beaumont Hague Cedex, France (Received September 16, 1999) The analytical protocols currently used to measure plutonium in urinary excretion consist of radiochemical purification, the source preparation by electroplating and alpha spectrometry measurements. Such a procedure is of limited relevance to the individual monitoring of workers exposed to plutonium and mixed oxides. The use of ICP-MS, which takes much shorter counting time, was investigated with the aim of assessing the capabilities of this analytical tool for the determination of urinary excretion of plutonium. It has been shown that the detection level of 1 mbq. l 1 could be achieved after a radiochemical purification process for a standard ICP-MS set -up. Introduction Individuals, occupationally exposed to plutonium, or mixed uranium and plutonium oxides, are monitored for internal contamination by measurements of urinary excretions. These compounds are known for their low transferability so the fraction of plutonium excreted in urine is very low, hence highly sensitive analytical techniques are necessary to achieve an accurate determination. The routinely-used protocol is divided into two parts, i.e. chemical purification and analytical measurement. The first stage consists of selective separation of each actinide from the urine. The second stage uses alpha spectrometry to reach activity levels of around 1 mbq. l 1. Such sensitivity is necessary to meet the requirement of individual monitoring in accordance with ICRP recommendations. 1,2 Lower detection limits have been reported, but with longer counting times than can be used in routine monitoring. The aim of this work was to assess the conditions in which Inductively Coupled Plasma Mass Spectrometry (ICP-MS) could be used to obtain similar detection levels for plutonium but in shorter time periods. ICP-MS has been shown to be a suitable technique for measuring urinary excretions of uranium, 3 5 even using very simple dilution protocols. 3,4 Since plutonium isotopes have much higher specific activities than most uranium isotopes, * the applicability of ICP-MS for the measurement of Pu is less attractive. Therefore, the direct dilution approach was not investigated for plutonium as an ICP-MS measurement protocol. Measurement of 238 Pu by ICP-MS is not desirable 1, * 234 U:2.30. 10 8 Bq. g 1, 235 U:8.00. 10 4 Bq. g 1, 238 U:1.24. 10 4 Bq. g 238 Pu:6.33. 10 11 Bq. g 1, 239 Pu:2.29. 10 9 Bq. g 1, 240 Pu:8.39. 10 9 Bq. g 1, 241 Pu:3.81. 10 12 Bq. g 1, 242 Pu:1.46. 10 8 Bq. g 1. because for a sensitivity of 150,000 cps/µg. l 1 and a background level of 40 cps at m/z = 238, currently observed when uranium measurements are made, the baseline corresponds to a concentration of 169 Bq. l 1 for 238 Pu. This study deals with the measurement of 239 Pu, 240 Pu and 242 Pu by ICP-MS after a chemical purification process. A conventional purification protocol and a highly sensitive configuration for ICP- MS set-up were chosen. The measurement limits for plutonium isotopes using ICP-MS were first estimated by using synthetic solutions for these conditions. Secondly, urine samples spiked with a known amount of plutonium were analysed by ICP-MS, in order to assess which concentration levels could actually be measured after chemical purification followed by preconcentration. The results obtained for plutonium are discussed, taking into account the specificity of the urine medium, to determine whether or not a sample introduction device for desolvating the sample is likely to actually enhance ICP-MS sensitivity. Based on these results, the use of ICP-MS, in the field of radiotoxicological analysis for plutonium determination in urine, is discussed. Plutonium spiked solutions Experimental Four synthetic solutions spiked with a known amount of 239 Pu were prepared. In these samples, hereafter referred to as S1, S2, S3 and S4, the 239 Pu activities were equal to 20, 10, 1 and 0.2 Bq. l 1, respectively. 0236 5731/2000/USD 17.00 Akadémiai Kiadó, Budapest 2000 Akadémiai Kiadó, Budapest Kluwer Academic Publishers, Dordrecht
The 239 Pu standard solution used to prepare them also contained 240 Pu as an impurity, which corresponds to 3.1% of the 239 Pu activity. Furthermore, a nitric acid solution containing 0.9 Bq. l 1 of pure 242 Pu, referred to as S T, was prepared to determine the response of the instrument at m/z = 242. From a pool of urine samples collected from unexposed people, separate fractions were sampled: two litres (U 1 and U 2 ) were spiked with 239 Pu and 242 Pu and one liter was used as blank (U Bl ). Urine samples U 1 and U 2 were prepared by weight. Each of them contained 9 mbq. l 1 of 242 Pu, and 80±7 mbq. l 1 and 10.0±1.3 mbq. l 1 of 239 Pu, respectively. These activities were checked by analysing synthetic solutions of identical volumes by alpha-spectrometry, using a conventional electrodeposition procedure (H 2 SO 4 0.02 v/v, ph = 2, I = 0.7 A and t = 3600 s) and ALADINtype alpha spectrometers containing implanted and passivated junction silicon detectors (Eurisys measures, Strasbourg, France). Purification protocols The radiochemical purification process involved a protocol set-up for routine monitoring purposes. 6 Typically, it is based on a one-litre urine sample to achieve concentration level of less than 1 mbq. l 1 after 3 days counting time of an electroplated source by alpha-spectrometry. Such protocol is currently used to analyse actinides in urine samples. The same procedure was used in this investigation to process the samples before ICP-MS measurements for urine samples and the urine blank. The chemical yield was determined by comparing the measured values for the tracer ( 242 Pu) obtained for the synthetic solution (S T ) and the purified plutonium solutions. ICP-MS measurements The following reagents, standards and blanks were used: ultrapure HNO 3 (Normatom, Prolabo, France), bismuth stock standard solutions 10 mg. l 1 and multielement stock standard solution containing depleted uranium 10 mg. l 1 (Spex, Ind, USA), and ultrapure water with a resistivity of less than 18 ΜΩ. cm. A PLASMAQUAD PQ 2+ (VG Elemental, Winsford, Cheschire, UK), equipped with a quadrupole mass spectrometer, was used in a configuration known as the S option. This configuration consists of adding an additional primary pump to upgrade the vacuum in the samp le-skimmer interface and enhance sensitivity. The sample was delivered using a peristaltic pump (Minipuls3, Gilson, Villiers le Bel, France) and a glass concentric nebulizer. The aerosol produced was directed through a water-cooled Scott spray chamber into a quartz plasma torch. Additional argon was supplied to the torch as a coolant and as an extra support for the plasma. The plasma was maintained at 1,350 W. Blank and standard solutions prepared with 2% HNO 3 were used to determine the contribution of the nitric acid in measurement of plutonium, as well as the sensitivity and the stability of the apparatus under normal conditions. In these experiments, the ICP-MS was calibrated using uranium standard solutions to avoid the handling of plutonium. The sensitivity levels at m/z = 238 and 239 were assumed to be the same, given that uranium and plutonium ionisation potentials are similar (around 7 kev). Prior to any measurement, all the solutions were spiked with a known amount of 209 Bi used as an internal standard to correct for any variation in the ICP-MS. The analytical procedure used to determine the amount of plutonium in urine consists of the following stages: (a) acid blank 2% HNO 3, (b) acid standard 2% HNO 3, (c) acid blank 2% HNO 3, (d) urine blank, (e) urine standard, (f) urine blank, (g) urine samples. Results and discussions In order to assess the sensitivity of the ICP-MS technique for plutonium analysis, a first set of measurements was made on synthetic solutions spiked with 239 Pu in order to determine the background and to optimise the counting time as a function of plutonium concentration. The next tests were performed to validate the use of 242 Pu as a tracer for ICP-MS measurements. Using these results, the sensitivity of ICP-MS could be assessed for the analysis of urines spiked with 239 Pu and 242 Pu after chemical treatment. For the results described below, quoted uncertainties are the standard deviation observed during the acquisition of the sample of interest for 209 Bi (internal standard) and for plutonium isotopes. To determine the overall uncertainty of the yield and the initial plutonium concentration, the statistical errors were combined using the conventional error propagation formula. Background The blank contribution at m/z = 239 was measured for HNO 3 2% and for a uranium standard solution (C = 1 µg. l 1 ). The difference between the results observed for the two sets of solutions at m/z = 239 is due to the 238 U 1 H generated in situ which is isobaric to 239 Pu. The uranium hydride abundance was determined, verifying that the background was the same at m/z equal to 239, 240 and 242 for the nitric acid solution (Table 1). 398
Table 1. Background determined at m/z = 239 Solution C U, µg. l 1 N 238 U, cps N [m/z = 239], cps N [m/z = 240], cps N [m/z = 242], cps HNO 3 2% 0 50 2 2 2 Uranium standard 1 150,000 6 2 2 Table 2. Comparison between the sensitivity (S) observed at m/z = 239, 240 and 242. The results given in this table have to be compared with the sensitivity at m/z = 238 (150,000 cps/ppb) Solution N 239 ± SD, S (m/z = 239), N 240 ± SD, S (m/z = 240), N 242 ± SD, S (m/z = 242), cps cps/ppb cps cps/ppb cps cps/ppb S1 1,350 ± 17 154,575 ± 1,947 13 ± 2 175,920 ± 27,064 no 242 Pu no 242 Pu S2 680 ± 10 155,720 ± 2,290 6 ± 1 162,387 ± 27,065 no 242 Pu no 242 Pu S3 64 ± 4 146,560 ± 9,160 N.D. N.D. no 242 Pu no 242 Pu S4 12 ± 2 137,400 ± 22,900 N.D. N.D. no 242 Pu no 242 Pu S T mes 1 no 239 Pu no 239 Pu no 240 Pu no 240 Pu 900 ± 20 146,920 ± 2,920 S T mes 2 no 239 Pu no 239 Pu no 240 Pu no 240 Pu 892 ± 18 144,702 ± 2,920 N.D. Not detected. For the uranium standard solution, the values measured at m/z = 240 and 242 were similar to those observed for 2% HNO 3. Therefore, the uranium hydride contribution ( 238 U 1 H) at m/z = 239 could be assessed by first determining the mean background (B, ) on the blank solution: N N N ( B ) = 3 [ m/ Z= 239] [ m/ Z= 240] [ m/ Z= 242] 2 cps The result was subtracted from the value measured at m/z = 239 using a 1 µg. l 1 uranium standard solution (Table 1), leading to a hydride contribution of 4 cps. The hydride level was estimated by dividing the hydride contribution by the measured values at m/z = 238 for the 1 µg. l 1 uranium standard solution. The hydride level equal to 2.7. 10 5 found here is consistent with that given by the manufacturer and that published in the literature. 7 For solutions containing both uranium and plutonium at trace level, the measurement of 239 Pu using, ICP-MS is, therefore, almost impossible due to the presence of uranium hydride. However, for the measurement of plutonium in urine, a chemical purification process is necessary to reach concentration of less than 1 mbq. l 1. Since each actinide is eluted in a separate fraction, hydride generation will be negligible. Counting time For ICP-MS measurements, during a 1-minute acquisition, numerous sweeps are made on the mass range in question, depending on the number of elements to be measured. The dwell time corresponds to the counting time per sweep for a given element. When several elements are measured during the same acquisition, different dwell times can be chosen in order to reach sufficient stability for each element. As a general rule, the longer the dwell time, the better the stability. However, if the dwell time selected for one element is too long, it may significantly deteriorate the stability of the measurement for the second element, even when this one is present in much higher concentration such as the internal standard. Therefore, the dwell time was optimised using synthetic solutions. As the results are normalised to the internal standard variation, 209 Bi, present in sufficient concentration (1 µg. l 1 ), the dwell time on the corresponding mass was kept to 2 ms. An identical dwell time was chosen for all plutonium isotopes. The influence of this dwell time was studied in a range of 10 to 500 ms. Increasing the dwell time led to improved stability, whereas no influence on sensitivity was observed. For this study, a dwell time for the plutonium isotopes of 100 ms was chosen in order to ensure good stability while having a negligible effect on the stability of the bismuth. Sensitivity Preliminary tests were carried out by tuning the analytical tool at m/z = 238, using uranium standard solutions. The compliance between the uranium sensitivity and that observed for 3 different plutonium isotopes, at m/z = 239, 240 and 242 based on the introduced activity and their specific activity, was checked (Table 2). As the sensitivity observed for the plutonium isotopes is the same as at m/z = 238, the uranium standard solution could be used to tune the machine prior to plutonium measurement. The best sensitivity that could be obtained by using ICP-MS was quantified for a synthetic solution. For a 239 Pu concentration of 200 mbq. l 1, a peak with a net surface of 12 cps was measured. Since the background at the corresponding mass is equal to 2 cps, a concentration 399
of 100 mbq. l 1 should be detectable. Knowing the physical limits of the analytical tool for determining the concentration level of the initial urine sample, a radiochemical purification process was performed for different urine samples and the preconcentration factor was assessed. Urine Prior to the measurement of Pu in urine, the response for 242 Pu used as a tracer was determined in the operating conditions using S T solution (Table 2). Furthermore, urine blank measurement showed no background due to the matrix. A one-litre urine sample, corresponding close to the daily excretion, was treated. The eluted recovering solution, containing only the plutonium separated from the other actinides and urine components, was made of 20 ml of hydroxylamine hydrochloride (7 g. l 1 ). This solution was evaporated and dissolved in 10 ml ultrapure 0.5M HNO 3, to perform ICP-MS analysis, leading to a concentration factor of 100. The yields determined for both urine samples (Table 3) are consistent. Moreover, the yields are close to currently reported values, obtained using this kind of purification procedure, confirming the suitability of the 242 Pu isotope as a tracer. The different urine samples were also spiked with 239 Pu and were measured simultaneously using 242 Pu isotope. The chemical yield and the measured values in cps at m/z = 239 (400 cps for U 1 and 60 cps for U 2 ) allow to calculate the initial concentration (Table 4). For both activity levels, the observed values and the spiked values were fairly consistent. Table 3. Chemical yield (R) determined for 242 Pu measurement using ICP -MS Sample N ± SD, cps N st ± SD, cps R ± R, % U1 677 ± 17 896 ± 24 75.6 ± 6.7 U2 710 ± 16 896 ± 24 79.2 ± 5.6 Table 4. Values of the initial concentration (C 0m, 239 ) determined after purification Sample C 0m, 239 ± C0m, 239, R ± R, C0m, 239 ± C0m, 239, mbq. l 1 % mbq. l 1 U1 63.3 ± 2.4 75.6 ± 6.7 83.8 ± 8.0 U2 8.2 ± 0.3 79.2 ± 5.6 10.4 ± 0.8 Measurements of plutonium using ICP-MS are limited and closely related to the half-lives of the different plutonium isotopes. ICP-MS results are given in g. l 1 which imply that only the atoms present in the solution are detected. Then, fixing the concentration to be assessed in Bq. l 1 (1 mbq. l 1 ), the higher the specific activity, the lower the equivalent concentration to be detected in g. l 1. Thus, unlike uranium and thorium that could be measured by ICP-MS after direct dilution, due to their long half-lives, a complete purification protocol is necessary to measure urinary plutonium. Using a high sensitivity configuration ICP-MS (PQ 2+ VG elemental), when a chemical purification was carried out on a onelitre urine sample containing 10 mbq of 239 Pu with a chemical yield around 80%, the measured value at m/z = 239, was 60 cps. Knowing that 6cps could be measured (Table 2), a factor of 10 could be achieved, making it possible to detect concentrations as low as 1 mbq. l 1 using the high sensitivity configuration. However, to achieve such a concentration level, the counting time will be drastically reduced from 3 days to 10 minutes. The lower detectable concentration still remains equal or higher to that required, i.e., 1 mbq. l 1. To obtain such detection levels, some technical options might be helpful: (1) the screen torch which may allow the sensitivity to be increased by a factor of two, (this gain being assessed by the manufacturer) and (2) the MISTRAL which makes it possible to gain a factor of ten in sensitivity on the actinide mass range. 7 Moreover, since an eluted volume of 10 ml (leading to a concentration factor of 100) makes it possible to run two measurement trials, each composed of ten acquisitions of one minute, its reduction might allow to improve the detection level of plutonium. The overall gain to be expected from combining these options will be further investigated. Since the use of the MISTRAL implies a higher sample consumption and is more complicated to set up, this option will be avoided where possible. Conclusions The present study confirms that ICP-MS could be used to perform plutonium measurements in a relatively simple configuration to achieve detection level of 1 mbq. l 1. The ways to increase sensitivity are clearly identified and have to be further investigated in order to confirm the measurement protocol which is considered to be the most relevant for monitoring the workers, i.e. the screen torch and the reduction of the eluted volume. 400
The optimisation of these two parameters might allow to reach activity levels of less than 1 mbq. l 1 for 239 Pu. Furthermore, the use of a single tracer will simplify the comparison between the two analytical techniques, alpha spectrometry and ICP-MS. As far as counting times are concerned, ICP-MS could be an interesting alternative to alpha spectrometry to reach these concentration levels. References 1. International Commission on Radiological Protection (ICRP). Human Respiratory Tract Model for Radiological Protection, ICRP Publication 66, Oxford, 1993. 2. International Commission on Radiological Protection (ICRP), Individual Monitoring for Internal Exposure of Workers. ICRP Publication, 78, Oxford, 1997. 3. Z. KARPAS, L. HALICZ, J. ROIZ, R. MARKO, E. KATORZA, A. LORBER, Z. GOLDBART, Health Phys., 71 (1996) 879. 4. N. BAGLAN, C. COSSONNET, F. TROMPIER, J. RITT, P. BERARD, Implementation of ICP -MS protocols for uranium urinary measurements in worker monitoring, submitted to Health Phys. 5. Z. KARPAS, A. LORBER, E. ELISH, P. MARCUS, J. ROIZ, R. MARKO, R. KOL, D. BRIKNER, L. HALICZ, Health Phys., 74 (1998) 86. 6. J.C. HARDUIN, B. PELEAU, D. LEVAVASSEUR, Radioprotection, 31 (1996) 229. 7. R. CHIAPPINI, J.M. TAILLADE, S. BREBION, J. Anal. Atomic Spectrom. 11 (1996) 497. 401