Individual monitoring of workers exposed to the risk of intake of actinides requires suitable methods for measuring low level of excreted activity. The current protocols used for actinides analysis in bioassay are usually complicated and highly time consuming. In this work, a protocol based on the microwave digestion of urines followed by the separation of actinides using calix[6]arene-based chromatography columns and their measurement by a quadrupole ICP-MS is developed and validated, for the first time, on urine samples containing the three actinides, U, Pu and Am. With this protocol, the total analysis time is about 2 days, including the mineralization of urine and the chromatographic separation of actinides. Detection limits of actinides in urine are determined and compared to those obtained after “dilute and shoot” ICP-MS analysis or after alpha spectrometry measurement.
The analysis of actinides is of great importance due to their high radiotoxicity and the severe damages their alpha radiation may induce at the cellular level. After contamination, radiotoxicological analyses of bioassay samples (e.g. urine and feces) are required to detect and quantify the contamination by the actinides and to estimate internal dose related to this actinides intake in the body. The most widely used technique for determination of alpha emitters (U, Pu, Am) in biological samples is alpha spectrometry. Alpha particles heavily ionize matter and then quickly lose their kinetic energy. Therefore alpha particles deposit all their energies along their short paths in matter. To detect alpha particles emitted by actinides in urine or feces, it is necessary to destroy all the organic matter and to separate the actinides from mineral compounds present in bioassay samples. Furthermore, the measurement of actinides by alpha spectrometry requires the separation of actinides from each other because of their interferences in alpha spectrometry. Therefore the current methods using alpha spectrometry to detect contamination by actinides require mineralization of bioassay samples then separation of actinides on various chromatographic columns and finally long count- ing times to achieve very low activity level (0.5 mBq/L) [
The aim of this work was to optimize all steps from the mineralization of urine samples with microwave oven to the actinides separation with calixarene columns on urine samples containing the three actinides U, Pu and Am and to evaluate the robustness of the optimized protocol on various urine samples. In this paper, the ICP-MS measurement was carried out for the detection of actinides for convenient purposes, but this separation protocol has been developed to be suitable for alpha spectrometry measurement, and the advantages and the drawbacks of both techniques for actinides analysis are also discussed.
239Pu and 243Am were purchased from Eckert & Ziegler (USA) and AEA Technology (Isotrak QSA Amersham, USA), respectively. Natural uranium was from SPEX Certiprep, Ind. (USA). All chemicals from Aldrich or VWR used in this study were of analytical grade and the water was purified by a Milli-Q® Synergy 185 water purification system from Millipore (Merck). Calixarene molecules were synthesized as described in the patent [
Digestion of urine samples were performed with a closed vessel digestion system, the Milestone EthosOne© microwave (Thermo Electron, France) and the analyses of total organic carbon with a Vario TOC cube analyser (Elementar, France).
All the measurements of 239Pu, 243Am and 238U in aqueous phases were performed by Inductively Coupled Mass Spectrometry (ICP-MS) using a quadru- pole mass spectrometer “Agilent 7700” (Agilent, France).
In case of suspected contamination of nuclear workers by actinides, their urines are collected on 24 hours (about 1 L) and analyzed to detect and quantify the contamination. To control the process of urine treatment (mineralization, chromatographic separation) before the actinides measurement, each urine sample is spiked with an isotope of each actinide not initially present in the urine such as, for example, 233U, 242Pu and 243Am.
In our study, urine samples (1000 mL) from people non-exposed to actinides contamination were collected then spiked with 0.1 µg (or 2.5 mBq) of natural uranium, 0.1 µg (or 230 Bq) of 239Pu and 0.3 ng (or 2.1 Bq) of 243Am, to control the contamination. After acidifying with 20 mL of 8 M HNO3, the urine was heated under reflux on a hot plate with magnetic stirring for 2 h. In the still warm urine, ammonia water (20 wt%) was added until precipitation of alkali-earth phosphate was observed. The supernatant was discarded and the precipitate was centrifuged and digested by microwave oven before actinides separation with extraction chromatography. In this paper, the microwave digestion was optimized on urine samples spiked with natural uranium. The mineralization efficiency was checked by controlling the extraction yield of uranium from mineralized urine samples on the carboxylic calixarene column (CC column) and comparing it with those obtained after a classical mineralization of urine samples using an ashing step of the alkali-earth phosphate precipitate in a muffle furnace at 1170 K during about 14 h. The influence of nitric acid volume (10 to 50 mL of 67 wt% nitric acid) and hydrogen peroxide volume (2 to 10 mL of 30 wt% hydrogen peroxide) and the effect of the heating profile were investigated. This digested solution is called “mineralized urine”.
The resin CH used in the hydroxamic calixarene column is composed of 1.78 × 10−5 mol of 1, 3, 5-OCH3-2, 4, 6-OCH2CONHOH-p-tert-butylcalix[
The pH of mineralized urine is increased, using ammonia water, to about 2.7 before its loading on the CH column for Pu extraction. pH values of 3 and 3.5 were also tested for Pu extraction. The CH column was then rinsed with 10 mL of 0.04 M NaNO3 solution adjusted at the same pH than the mineralized urine and the combined effluent (from the loading and the rinsing steps) was retained for uranium and americium separation (solution A). Plutonium was finally eluted with 30 mL of hydrochloric acid solution (2, 3 or 4 M) or with 30 mL of 0.1 M hydroxylammonium chloride in 2 M HCl (Pu fraction). The pH of the retained americium and uranium fraction (solution A) is adjusted at 5.2 ± 0.1 using ammonia water, before its loading on the CC column. The CC column was then rinsed with 10 mL of 0.04 M NaNO3 solution adjusted at pH 6.0 ± 0.1. In this paper, two eluents, ethylene diamine tetraacetic acid (EDTA) and diethylene triamine pentaacetic acid (DTPA) were investigated for the separation of americium/uranium on the CC column. The effect of acetate buffer in these eluents was also studied. Finally, the elution of uranium is performed with 30 mL of 1 M HCl. The optimized protocol is illustrated in
All chromatographic separation steps were performed at room temperature (22˚C ± 4˚C) to ease the use of the separation protocol.
Aliquots are sampled before and after the extraction of actinides on CH and CC columns and then analyzed by ICP-MS to determine the extraction yield for each actinide. Each elution fraction was also analyzed by ICP-MS to determine the elution yield (YE) of actinides and the chemical recovery (R) of each actinide, according to the equations (1) and (2), respectively. Each sample was prepared using a 2-fold dilution in nitric acid (2 wt%) before its ICP-MS measurement.
where Qurine, Qextracted, and Qeluted is the quantity of each actinide initially in the urine sample, extracted and eluted from the calixarene column CC or CH, respectively. [An]E, [An]0 and [An]L+R is the concentration of each actinide An (An = 238U, 239Pu or 243Am) in the elution fraction, in the urine and in the combined effluent (loading + rinsing steps), respectively. VE, Vurine and VL+R is the volume of the elution fraction, the volume of urine, the volume of the combined effluent (loading + rinsing steps), respectively.
All analytical results in this paper are from at least three replicates, and the uncertainties are the standard deviations of the replicates expressed with k = 1. In all these experiments, the mass balance for each actinide was checked.
As mentioned earlier, the mineralization of urine is essential for detection of alpha articles emitted by the actinides. The aim of this study was to optimize this mineralization step. The classical mineralization was usually performed by calcination of the calcium phosphate precipitate containing the actinides in a muffle furnace. This step takes about 14 hours. The microwave technology has been already used for mineralization of biological samples [
The required quantity of reagents, nitric acid and hydrogen peroxide, two usual reagents used for mineralization [
When the quantity of nitric acid increases, a more efficient mineralization is achieved leading to a better extraction of uranium on CC column. But these results demonstrate that nitric acid cannot be used alone for a complete digestion of urine sample, as the uranium extraction yield on the CC column is still lower than those achieved after classical mineralization (79 ± 6)%.
Then a second series of experiments was performed by adding various volumes of hydrogen peroxide (2 to 10 mL of 30 wt% hydrogen peroxide) to 30 mL of nitric acid (67 wt%), as it is known that the use of hydrogen peroxide enhances the oxidation properties of nitric acid especially in the digestion of organics [
Various heating profiles were tested with one or two-step temperature ramp as
described in
To complete this study, the analyses of total organic carbon were performed on urine samples before and after mineralization with the optimized procedure using the microwave program P4. The results show that more than 99.5% of organic carbon amount is destroyed thanks to this mineralization procedure.
Finally this mineralization procedure was also validated on the actinides Pu and Am in mixture in urine and their extraction was then performed on CH and CC column, respectively. The extraction yields were (78 ± 15)% for Pu and (92 ± 2)% for Am, respectively. All these results allow validating the mineralization protocol of urines by microwave digestion for actinides (U, Pu and Am) analysis, with a total recovery of actinides in mineralized urine.
In conclusion, the goal of this study on mineralization step was achieved: it was demonstrated that the protocol of mineralization with microwave furnace exhibits excellent digestion performance. With the judicious choice of reagents and heating profile allowing total sample decomposition, shorter mineralization time (about 1.5 h) is achieved compared to classical calcination in muffle furnace (14 h) with a high sample throughput, since up to 10 urine samples can be mineralized per run.
The previous studies on hydroxamic calix[
Program (total mineralization time) | Heating profile | U extraction (%) | |
---|---|---|---|
Time (min) | Temperature (˚C) | ||
P1 (2h05) | 0 à 5 5 à 10 10 à 20 20 à 35 | 0 à 100 100 100 à 200 200 | 53 ± 16 |
P2 (1h35) | 0 à 10 10 à 25 25 à 35 35 à 65 | 0 à 100 100 100 à 200 200 | 65 ± 7 |
P3 (1h40) | 0 à 5 5 à 25 | 0 à 200 200 | 63 ± 13 |
P4 (1h25) | 0 à 10 10 à 40 | 0 à 200 200 | 62 ± 7 |
arene extractants showed that the extraction of actinides by these molecules is controlled by the deprotonation of their hydroxamic or carboxylic functions and thus is pH dependent [
In this paper, the influence of the extraction pH on the separation of Pu from U and Am was investigated with CH column. These experiments were carried out on urine samples containing the mixture of actinides, U, Pu and Am. Three pH values were studied: 2.7, 3 and 3.5, because previous results showed that this pH range can be used to separate Pu from U and Am. The results are presented in
As we can see, the increase of extraction pH improves the extraction of plutonium on CH column, as it promotes the deprotonation of hydroxamic functions of calix[
In conclusion if the separation of actinides is critical (for example in case of alpha spectrometry measurement), it is recommended to perform the first extraction step on CH column at pH 2.7 to achieve the best separation of Pu from U and Am. Otherwise, it may be judicious to carry out the plutonium extraction at pH 3 to improve its extraction on CH column.
To develop a protocol that could be suitable also for alpha spectrometry measurement, hydrochloric acid and hydroxylammonium chloride were investigated as eluent, since these reagents are commonly used in protocols for plutonium determination [
Extraction pH | 2.7 | 3 | 3.5 |
---|---|---|---|
Pu extraction (%) | 81 ± 9 | 97 ± 1 | 91 ± 2 |
U extraction (%) | 3 ± 4 | 10 ± 4 | 25 ± 1 |
Am extraction (%) | 3 ± 1 | 0 ± 0 | 11 ± 3 |
Eluent | 2 M HCl | 3 M HCl | 4 M HCl | 0.1 M hydroxylammonium chloride in 2 M HCl |
---|---|---|---|---|
Pu elution (%) | 62 ± 9 | 62 ± 7 | 76 ± 6 | 78 ± 8 |
The best Pu elution is achieved with 4 M HCl and 0.1 M hydroxylammonium chloride in 2 M HCl. This last reagent was chosen since it is already used in other protocols for Pu elution.
A previous study demonstrated that the use of 0.1 mM EDTA solution at pH 3.2 enhances the U/Am separation after their co-extraction at pH 5.2 on carboxylic calixarene column [
In a first time, the repeatability of U/Am separation from carboxylic calixa- rene column was investigated with 0.1 mM EDTA solution at pH 3.2, as suggested in our previous work, on four urine samples. The results exhibit that the U/Am separation is not repeatable from sample to sample in these conditions (data not shown).
As the extraction of actinides by the calixarene columns is pH-dependent [
pH value starts at 5.2, that is the pH used for the extraction and rinsing steps, and then it increases quickly until about 6. This pH increase goes along with the beginning of the americium elution. The pH is relatively stable until that 25 mL of EDTA solution is percolated through the column and then decreases to achieve the final pH of 3.2. This decrease of pH is more or less rapid and the uranium elution starts to be significant as the pH is below 5. The correlation between the pH of EDTA solution measured at the bottom of the column and the elution yields of U and Am is illustrated in
The actinides elution from the carboxylic column is controlled by two reactions:
・ The reprotonation of the carboxylic functions of calixarene molecules when the pH of mobile phase decreases in the calixarene column
・ The action of chelating agent, as EDTA (ethylene diamine tetraacetic acid), that can complex the actinides and then lead to their elution from the calixarene column.
The complexation constant of americium by EDTA is much higher than the one between uranium and EDTA, as shown in
These new results on various urine samples show that the pH of EDTA solution has to be controlled to achieve a reproducible separation of U/Am. The use of buffering agent in EDTA solution was then investigated.
New experiments were carried out on urine samples containing a mixture of U and Am. For U/Am separation, 0.1 mM EDTA solution with 0.03 M sodium acetate as buffering agent was used to control the pH during the elution step. pH of 4, 5.2 and 6 were investigated as buffer pH values. The elution yields for U and Am were determined and the pH of EDTA eluent was measured at the bottom of the column in aliquots of 10 mL. The results for pH 4 or 5.2 are not presented, because no separation of U/Am was achieved in these conditions, the co-elution of both actinides started with the beginning of the loading of the mixture of EDTA solution with sodium acetate buffer into the CC column. The results obtained for EDTA solution with sodium acetate buffer at pH 6 are shown in
The results demonstrate that the use of sodium acetate buffer in EDTA solution leads to a better control of pH during the elution step and then prevents the co-elution of uranium with americium. However, it is noticed that the Am elution yield is slightly lower compared with EDTA eluent without buffering agent (about 70% compared to 100%). This could be explained by a salt effect due to the increase of ionic strength, along with the addition of sodium acetate buffer in the EDTA eluent [
To improve the americium elution yield, the influence of EDTA concentration in presence of sodium acetate buffer (pH 6) on the separation of U/Am from the CC column was investigated on urine samples after mineralization. The results are summarized in
When the EDTA concentration increases, the americium elution yield increases but the co-elution of uranium also increases, driving by the actinides complexation by EDTA. Thus this parameter cannot be used to improve the elution of americium, because EDTA is not a chelating agent selective enough for the U/Am separation.
Log K1 (I = 0) | Am3+ | |
---|---|---|
EDTA | 13.7 | 19.7 |
DTPA | 11.0 | 26.2 |
EDTA | DTPA | ||||
---|---|---|---|---|---|
0.1 mM | 0.5 mM | 1 mM | 0.01 mM | 0.05 mM | |
Eluted Am (%) | 72 ± 4 | 71 ± 7 | 95 ± 6 | 70 ± 7 | 81 ± 8 |
Eluted U (%) | 2 ± 2 | 25 ± 3 | 49 ± 5 | 1 ± 1 | 3 ± 2 |
To optimize the americium elution, the use of DTPA (diethylene triamine pentaacetic acid) as eluent instead of EDTA was studied. Indeed, according to the complexation constants of EDTA and DTPA for uranium and americium given in
Two concentrations of DTPA were examined: 0.01 mM and 0.05 mM. The pH of DTPA solution was controlled with 0.03 M sodium acetate buffer at pH 6. The results are presented in
To validate this protocol, a set of experiments were carried out on 10 various urines samples (1000 mL) spiked with a mixture of actinides U, Pu and Am according to the optimized protocol using calixarene columns illustrated in
These results show the robustness of this protocol for actinides separation. Very good extraction yields of plutonium on the hydroxamic calixarene column and of uranium and americium on the carboxylic calixarene column are achieved with a good repeatability that confirms the good affinity of these extractants for actinides from urine samples. The first calixarene column allows an efficient separation of Pu from U and Am. By this way, the possible polyatomic interference between 1H238U and 239Pu would not be a concern for ICP-MS measurement with a quadrupole mass spectrometer. Furthermore the measurement of both isotopes 239Pu and 240Pu is possible by ICP-MS, whereas they can’t be distinguished by alpha spectrometry measurement due to their alpha energy that are very close (<15 keV). However, the use of quadrupole ICP-MS to detect a possible internal contamination of 238Pu in urine is still impossible due to the short half-life of this isotope and the isobaric interference with 238U that is naturally present in all chemical reagents. The extraction of Pu on the second calixarene column is also very low leading to a very low Pu contamination in the americium fraction and in the uranium fraction. Concerning the separation of U from Am with the carboxylic calixarene column, the first elution step with DTPA solution in sodium acetate buffer allows a good recovery of americium (65%) with less than 10% of uranium. Then the last elution step with 1 M HCl allows the recovery of 55% of uranium with about 10% of americium. This poor separation of U/Am could be a concern for alpha spectrometry measurement of actinides due to possible
Hydroxamic calixarene column (CH) | Carboxylic calixarene column (CC) | |||||
---|---|---|---|---|---|---|
Pu | U | Am | Pu | U | Am | |
Extraction (%) | 87 ± 11 | 3 ± 2 | 2 ± 2 | 3 ± 3 | 92 ± 4 | 97 ± 4 |
Recovery (%) in Pu fraction | 80 ± 14 | <4 | <4 | - | - | - |
Recovery (%) in Am fraction | - | - | - | <2 | 8 ± 6 | 65 ± 11 |
Recovery (%) in U fraction | - | - | - | <3 | 55 ± 17 | 11 ± 8 |
interferences between 232U and 241Am or 243Am isotopes, when the resolution of alpha spectrometry measurement is not sufficient (>30 keV). But for ICP-MS measurement, this separation could be sufficient since there is no possible interference between uranium and americium. Furthermore the low recovery for uranium can be compensated by the high sensitivity of ICP-MS measurement for this element.
With this protocol, the total analysis time is about 2 days including the mineralization and separation steps and ICP-MS measurement, and is about 5 days if alpha spectrometry measurement is used. By considering the chemical recoveries of actinides with this separation protocol, detection limits that can be achieved with ICP-MS can be calculated from nitric acid blank measurement. These detections limits are summarized in
These results show that this separation protocol based on calixarene columns allows to improve detection limits by a factor 150 for U isotopes and 239Pu and by a factor 65 for 241Am, as compared with the “dilute and shoot” ICP-MS analysis of urine. It should be mentioned that the detection limits obtained for U isotopes are probably slightly overestimated because of the presence of natural uranium in all chemical reagents. Thus “blank” experiments (with all steps including mineralization, separation on calixarene columns) should be carried out for a better estimation of detection limits for U isotopes. Nevertheless, ICP-MS is the suitable technique for U measurements in urine and is recommended instead of alpha spectrometry [
For 239Pu and 240Pu, detection limits in urine achieved by ICP-MS measurement after actinides separation on calixarene columns are close to those obtained by alpha spectrometry measurement with “emergency protocol” [
DL (mBq∙L−1) | 234U | 235U | 238U | 239Pu | 240Pu | 241Am |
---|---|---|---|---|---|---|
“Dilute and shoot” ICP-MS analysis | 300 | 0.15 | 0.05 | 2500 | 10000 | 150000 |
ICP-MS after chemical separation | 2 | 8 × 10−4 | 3 × 10−4 | 20 | 60 | 2300 |
ICP-MS instrument, more efficient sample introduction system or the use of a sector field ICP-MS instead of a quadrupole one, a better sensitivity and a lower background signal can be achieved implying lower detection limits [
Rapid and reliable bioassays methods are in great demand for occupational radiation exposure monitoring of nuclear energy workers. Current protocols used for actinides determination in urine require lengthy sample preparation and long counting time, whereas protocols dedicated to radiological/nuclear emergency often exhibit higher detection limits. In this work, a protocol using a microwave digestion of urine samples followed by the separation of actinides using calix[
Bouvier-Capely, C., Legrand, A., Sylvain, A., Manoury, A. and Rebière, F. (2017) Operational Protocol for Detection of Contamination by Acti- nides U, Pu and Am in Urine Using Calixarene Columns: From Mineralization to Icp-Ms Measurement. American Journal of Analytical Chemistry, 8, 317-333. https://doi.org/10.4236/ajac.2017.85024