Analytica Chimica Acta xxx (2016) 1e7
Contents lists available at ScienceDirect
Analytica Chimica Acta journal homepage: www.elsevier.com/locate/aca
Determination of 236U in environmental samples by single extraction chromatography coupled to triple-quadrupole inductively coupled plasma-mass spectrometry Guosheng Yang a, b, c, Hirofumi Tazoe a, Masatoshi Yamada a, * a b c
Department of Radiation Chemistry, Institute of Radiation Emergency Medicine, Hirosaki University, 66-1 Hon-cho, Hirosaki, Aomori, 036-8564, Japan Division of Nuclear Technology and Applications, Institute of High Energy Physics, Chinese Academy of Sciences, China Beijing Engineering Research Center of Radiographic Techniques and Equipment, Beijing, 100049, China
h i g h l i g h t s
g r a p h i c a l a b s t r a c t
A simple 236U/238U analytical method has been developed. The separation required just one DGA column chromatography. 236U/238U atom ratios in soil were measured by ICP-MS/MS. 236U/238U atom ratios of (1.50 e13.5) 108 were observed in Japanese samples. activities of (2.25 236U e14.1) 102 mBq kg1 were found in Japanese samples.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 1 August 2016 Received in revised form 8 September 2016 Accepted 23 September 2016 Available online xxx
In order to measure trace 236U and 236U/238U in environmental samples with a high matrix effect, a novel and simple method was developed that makes the digestion and purification procedures compatible with advanced triple-quadrupole inductively coupled plasma-mass spectrometry. A total dissolution of sample with HF þ HNO3 þ HClO4 was followed by chromatographic separation with a single resin column containing normal type DGA resin (N,N,N0 ,N’-tetra-n-octyldiglycolamide) as the extractant system. The analytical accuracy and precision of 236U/238U ratios, measured as 236U16Oþ/238U16Oþ, were examined by using the reference materials IAEA-135, IAEA-385, IAEA-447, and JSAC 0471. The low method detection limit (3.50 106 Bq kg1) makes it possible to perform routine monitoring of environmental 236 U due to global fallout combined with the Fukushima Daiichi Nuclear Power Plant accident fallout (>105 Bq kg1). Finally, the developed method was successfully applied to measure 236U/238U ratios and 236 U activities in soil samples contaminated by the accident. The low 236U/238U atom ratios ((1.50 e13.5) 108) and 236U activities ((2.25e14.1) 102 mBq kg1) indicate 236U contamination was mainly derived from global fallout in the examined samples. © 2016 Elsevier B.V. All rights reserved.
Keywords: 236 U activity 236 U/238U Triple-quadrupole inductively coupled plasma-mass spectrometry Soil Fukushima Daiichi nuclear power plant accident
1. Introduction Naturally and artificially occurring 234U, 235U, and 238U have been extensively studied in previous studies [1e5]. However, the * Corresponding author. E-mail address: [email protected] (M. Yamada).
evidence for U source identification from 235U and 238U is limited, especially in soil samples, because the anthropogenic U has been strongly diluted by the large amount of natural U [3]. 236U has an extremely low natural abundance (236U/238U: ~1014 atoms) in the general environment [4,5]. The contribution from a decay of 240Pu in global fallout is also negligible compared to the directly
http://dx.doi.org/10.1016/j.aca.2016.09.033 0003-2670/© 2016 Elsevier B.V. All rights reserved.
Please cite this article in press as: G. Yang, et al., Determination of 236U in environmental samples by single extraction chromatography coupled to triple-quadrupole inductively coupled plasma-mass spectrometry, Analytica Chimica Acta (2016), http://dx.doi.org/10.1016/j.aca.2016.09.033
2
G. Yang et al. / Analytica Chimica Acta xxx (2016) 1e7
introduced artificial 236U since 240Pu has a long half-life (6564 y) and it is at a low concentration level in the environment. Therefore, presently, the ultra-trace amount of 236U found in environmental and biological samples has been mainly introduced by nuclear weapon explosions (global fallout), nuclear power generation and reprocessing of its fuel, and nuclear reactor accidents. The 236U/238U atom ratio varies with reactor, weapon, and fuel type [6e9]. Thus, 236 238 U/ U is potentially useful as a tracer for the source identification of anthropogenic U present in the environment. The Fukushima Daiichi Nuclear Power Plant (FDNPP) reactors used UO2 and mixed oxide ((U,Pu)O2) fuels. On March 11, 2011, the abundance of 235U in the Unit 1 reactor was estimated as 1.7%, and it was 1.9% for Unit 2 and 3 reactors, which are obviously different from the percent abundance of natural 235U (~0.72%) [2]. However, relatively high concentrations of natural U in Japanese soil (about 1e3 ppm) have prevented identification of the 235U/238U fingerprint from the FDNPP accident, especially for the granitoid areas in Fukushima Prefecture [1,10]. In addition, the 235U/238U atom ratios in plant samples obtained from Fukushima Prefecture after the FDNPP accident also can not reflect the release of U from this accident [2]. 236U concentration can provide additional information on nuclear activities in the contaminated areas, and the 236U/238U atom ratio can provide a far more sensitive tool for the detection of nuclear contamination with respect to other U atom ratios (e.g. 235 238 U/ U) [3,10,11]. Therefore, in this situation, 236U/238U is very appropriate and far more sensitive for demonstrating anthropogenic U contamination in the terrestrial environment from the FDNPP accident [12]. Although Sakaguchi et al. [13] measured 236U in river and sea water samples in the vicinity of the FDNPP, they could not evaluate the source. However, higher 236U/238U atom ratios up to ~106 were observed in airborne samples in the weeks immediately after the FDNPP accident [3]. Furthermore, the isotopic fractions of 236U isotopes in “black substances” collected along roadsides in Fukushima Prefecture revealed the substances contained U fallout from the FDNPP accident [10]. However, the 236U data in the environmental samples illustrating the U release from the plant have been limited until now, especially in soil. Thus, the characteristics of U as a refractory element that was emitted have not been fully clarified in the environment, even 5 years after the accident. The measurement of the trace 236U in the environment presents a challenge to the clarification. For the measurement of 236U, inductively coupled plasma-mass spectrometry (ICP-MS), thermal ionization mass spectrometry (TIMS) and accelerator mass spectrometry (AMS) have usually been employed. AMS is presently the method with the highest detection sensitivity for 236U. However, due to the high instrument price, there are only about 110 AMS facilities worldwide, and most of them are mainly applied to the routine analysis of 14C for dating purposes; only about ten can be used in the study of 236U [3,5,6,9e11,13e19]. TIMS is less preferred than ICP-MS because of greater difficulties in its sample preparation and measurement techniques. To date, ICP-MS is the most commonly applied mass spectrometric method for the investigation of actinides in the environment [20], and it has a detection limit for the 236U/238U atom ratio of about 108 [21]. In addition, until now, the only 236U data for the FDNPP accident were from the Vienna Environmental Research Accelerator (VERA) Laboratory at the University of Vienna, Austria, not from Japanese laboratories. Therefore, it is difficult to perform routine monitoring for 236U due to the FDNPP accident contamination, let alone to establish the 236U/238U database for future applications. However, from the viewpoint of risk assessment and protection against potential nuclear accidents, it is desirable to obtain sufficient 236U/238U background data, since more nuclear power plants and nuclear waste processing facilities
are being or will be built in the foreseeable future in the Northeast Asian region. Therefore, ICP-MS may be an alternative apart from AMS for routine monitoring of 236U. The ICP-MS analysis of 236U in environmental samples, however, remains a great challenge due to the matrix effect, the presence of polyatomic interferences such as 235UHþ, and the tailing effect from 238 U and 235U. Therefore, extraction chromatography, e.g. using UTEVA® and TRU resins, has been applied in tandem for the separation and purification of U from the remaining constituents of the matrix with a smaller solvent volume compared to ionic chromatography [22]. Among available extraction chromatographic resins, normal type DGA resin (N,N,N0 ,N0 -tetra-n-octyldiglycolamide as the extractant system) is another potential resin for U purification, but it requires further studies especially for that condition in which only a single DGA resin column is used [23]. In terms of the detection technique, collision/reaction cell technology was developed to overcome polyatomic interferences like 235UHþ and the tailing effect from 238U and 235U, since it is possible to get effective partitioning of ionemolecule reactions with O2 gas between Uþ and UHþ (Uþ þ O2 / UOþ þ O; UHþ þ O2 / UOþ þ OH) [24,25]. However, to apply collision or reaction cell technology with confidence, more environmental samples should be measured, including those with a high matrix effect. Therefore, an appropriate purification technique by DGA resin compatible with the confirmed collision/reaction cell technology for U analysis may be a potential solution for trace 236U analysis in high matrix effect samples. In the present study, a systematic analysis was made of normal dissolution (acid leaching) and total dissolution (complete digestion) procedures for U atom ratio measurements from environmental samples. Almost no loss of U was achieved during the DGA resin purification procedure. For triple-quadrupole inductively coupled plasma-mass spectrometry (ICP-MS/MS) analysis, an almost interference-free and non-peak tailing spectrum was obtained by measuring UOþ after adding O2 to an octopole collision/ reaction cell located between two quadrupole mass filters. The established method was first examined by measuring four soil and sediment reference materials, and second it was applied to the analysis of soil samples contaminated by the FDNPP accident. 2. Experimental section 2.1. Instrumentation An Agilent 8800 ICP-MS/MS (Agilent Technologies, Santa Clara, CA, USA) featuring an octopole collision/reaction cell situated between two quadrupole mass filters (first, Q1; second, Q2) was employed for analysis of U atom ratios. The introduction system was a high efficiency sample introduction system (APEX-Q, Elemental Scientific, Omaha, NE, USA) equipped with a PFA MicroFlow nebulizer. Flowing nitrogen (10 mL min1) was used to increase transport efficiency and signal stability. The sample uptake rate was ~0.5 mL min1. The ICP-MS/MS was optimized on a daily basis using 1 ng mL1 of SPEX-XSTC 622 standard solution containing U, and the optimized operation parameters are summarized in Table S1. Inductively coupled plasma-optical emission spectrometry (ICP-OES, SPECTRO BLUE TI, SPECTRO Analytical Instruments, Germany) was employed in the U purification procedure to monitor its major interfering elements. 2.2. Reagents and materials Analytical grade HNO3, HF, and HClO4 were obtained from Kanto Chemical Co., Inc. (Tokyo, Japan). Ultrapure grade HNO3 obtained from Tama Chemicals (Tokyo, Japan) and Milli-Q water (>18.2 MU cm1) were used for preparation of the final sample
Please cite this article in press as: G. Yang, et al., Determination of 236U in environmental samples by single extraction chromatography coupled to triple-quadrupole inductively coupled plasma-mass spectrometry, Analytica Chimica Acta (2016), http://dx.doi.org/10.1016/j.aca.2016.09.033
G. Yang et al. / Analytica Chimica Acta xxx (2016) 1e7
solution prior to ICP-MS/MS measurement. The 2 mL pre-packed cartridges containing DGA resin (particle size: 50e100 mm) were purchased from Eichrom Technologies LLC (Lisle, IL, USA). 10 mg L1 U standard stock solution in 5% HNO3 (SPEX-XSTC 622) was obtained from SPEX CertiPrep (Metuchen, NJ, USA). High purity O2 (>99.999%) was obtained from Taiyo Nippon Sanso Corp. (Tokyo, Japan). Marine sediment standard reference material (IAEA-135) was used to investigate the yield of every procedure. Other soil (IAEA-447 and JSAC 0471) and sediment (IAEA-385) standard reference materials were also applied to examine the method via measuring U atom ratios in samples with a high matrix effect. 2.3. U isotopes analysis About 1 g samples were ashed in a muffle oven at 450 C for 2 h to decompose organic matter. First, acid leaching (8 M HNO3 or concentrated HNO3) and total dissolution (HF þ HNO3 þ HClO4) procedures were evaluated as acid attack techniques for U dissolution using IAEA-135. Acid leaching was performed in PFA jars with lids (Savillex, Eden Prairie, MN, USA) on a hot plate at 180 C for 1 d. After filtration, the filtrate was evaporated to dryness. For the total dissolution, the residue was dissolved in 5 mL of 46% HF and 1 mL of 70% HClO4, and evaporated to near dryness again. This step was repeated five times more to remove remaining Si in the sample. After that, the residue was dissolved into 5 mL of 61% HNO3 and evaporated to dryness to remove the residual HF. For the normal dissolution, it is not necessary to perform Si and HF removal. After that, the sample was redissolved into 10 mL of 6 M HNO3 for chromatographic purification using the 2 mL prepacked cartridge containing DGA resin. Using 15 mL of 0.1 M HNO3, U was eluted from the resin, the eluent was evaporated to near dryness, and the residual was dissolved in 1.5 mL of 4% HNO3. For 238U concentration measurement via ICP-MS, a 20 mL aliquot was taken out and diluted with 4% HNO3 by a dilution factor of 2000 in order to get the sample concentrations within the calibration interval (0.01e1 ng mL1). The other fraction was analyzed for 236U/238U, 234 238 U/ U, and 235U/238U atom ratios via ICP-MS/MS. The signal intensities of 238U procedure blanks, incorporated during chemical treatment, were in the range of errors (1s for all data in this study) of the signal intensities of samples. Therefore, the digestion, purification, and dilution procedures did not introduce significant contamination for the U analysis. 3. Results and discussion 3.1. Total dissolution of U It is well known that anthropogenic Pu originating from global fallout can be recovered by acid leaching [20,26]. For soil and sediment samples, various kinds of acid leaching have been applied for 236U measurement previously. Some studies have demonstrated that 8 M HNO3 was sufficient for 236U leaching from soil samples [6,13,27e29]. Moreover, concentrated HNO3 was also used for 236U fallout leaching from soil matrix [5,8,30]. Since there is only a trace amount of 236U in environmental samples and potentially 236U may be present in the yield tracers, e.g. 232U and 233U, total dissolution and complete recovery during purification were desired for the accurate measurement of 236U with no addition of tracer. Furthermore, the aim of this study is 236U/238U atom ratio analysis for source identification studies in the future. To compare 236U/238U atom ratio among the presently studied samples and with other studies, it is better to release all of the U from the soil samples since 238 U also exists in the inner fractions in soil. A mineral acid mixture of HF with HNO3 or HCl was utilized for the total dissolution of 236U from soil and sediment samples [6,7,10,11,31e33]. In addition,
3
HClO4 and H3BO3 were applied synchronously to avoid the formation of insoluble fluoride. In order to get compatibility with the following DGA resin purification and ICP-MS/MS analysis, a systematic evaluation of normal dissolution (8 M or concentrated HNO3) and total dissolution (HF þ HNO3 þ HClO4) procedures for 236 U measurements from environmental samples was performed. For IAEA-135, screening of the results reported for the different U isotopes showed that there is a difference between the results obtained by using either the normal dissolution procedure or the total dissolution procedure [34]. For the normal dissolution procedure, 8 M HNO3 or concentrated HNO3 was considered to be an appropriate reagent for the recovery of global fallout 236U in soil and sediment samples, since this normal dissolution procedure was able to solubilize U held on mineral surfaces and in secondary phases (inner fractions), rather than attempting dissolution of 236U in primary minerals to avoid any natural U dilution [5,6,8,11,12,15,16,28e30]. Until now, the available 236U activities in the reference materials have been mostly deduced from 238U activities and 236U/238U atom ratios [7,9], as shown in Tables 1 and 2. Therefore, the yield of U isotopes during the digestion procedure was compared via 238U first. In the present study, for normal dissolution, concentrated HNO3 was a better acid for complete recovery of U isotopes in IAEA-135, since the measured 12.3 ± 1.1 Bq kg1 of 238U (n ¼ 3) by concentrated HNO3 leaching was in the range of the information values (11.3e18.1 Bq kg1) [34], as shown in Fig.S1. For total dissolution, the HF amount added in the digestion procedure had a significant effect on the recovery of U, as shown in Fig. 1. When 10 mL HF was added, the yield of 238U activity in IAEA-135 increased to 30.6 ± 3.6 Bq kg1 (n ¼ 3), which was in the range of the information values (27.0e36.5 Bq kg1) [34]. On the contrary, incomplete recovery of 238U for 5 mL HF addition (n ¼ 1) may be due to the insufficient acid amount, while the reason for incomplete recovery for more than 10 mL HF addition (n ¼ 1) was not clear. Finally, as shown in Tables 1 and 2, the 236U/238U atom ratio of (1.46 ± 0.35) 106 (n ¼ 3) was observed for total dissolution with 10 mL of 38% HF, 5 mL of 61% HNO3, and 3 mL of 70% HClO4. This value was in agreement with that given by Hotchkis et al. ((1.48 ± 0.37) 106) [9] and Srncik et al. ((1.48 ± 0.04) 106) [29]. However, a somewhat higher value of (2.0 ± 0.4) 106 was obtained by Lee et al. [7] In addition, 8.59 ± 2.61 mBq kg1 (n ¼ 3) of 236U was observed for IAEA-135, which was also somewhat lower than that by Lee et al. (11.2 ± 2.2 mBq kg1) [7]. The higher 236U activities deduced from 236 238 U/ U atom ratios by AMS and 238U activities by other analytical techniques, with different chemical and analytical procedures, may have less accuracy. In the same study [7], a relatively lower 236 238 U/ U atom ratio of (1.8 ± 0.5) 106 (n ¼ 3) and 236U activity of 10.3 ± 3.0 mBq kg1 were obtained by ICP-MS. The wide range of reported results indicates hot particles may be present in IAEA-135, which is believed to have some contamination from the Sellafield reprocessing plant or the 1957 Windscale accident [9]. As shown in Table 1, apart from 236U and 238U, 234U and 235U were also in the range of information values for IAEA-135. In brief, the total dissolution procedure used in the present study was sufficient to release U isotopes from environmental samples. 3.2. Chromatographic purification of U Since ICP-MS/MS and APEX-Q introduction systems are all susceptible to the matrix effect, it is necessary to carry out U purification first for the samples with a high matrix effect. It is well known that U can be separated from the matrix in environmental samples by extraction resins or anion exchange resins. In previous studies, TRU and UTEVA® resins were successfully applied for the separation and purification of U [4,5,7,9,11,16,27e30,33,35,36].
Please cite this article in press as: G. Yang, et al., Determination of 236U in environmental samples by single extraction chromatography coupled to triple-quadrupole inductively coupled plasma-mass spectrometry, Analytica Chimica Acta (2016), http://dx.doi.org/10.1016/j.aca.2016.09.033
4
G. Yang et al. / Analytica Chimica Acta xxx (2016) 1e7
Table 1 U isotopic information obtained for IAEA-135 after applying the total dissolution procedure. IAEA-135
Note
Value (Bq kg1)
No. of replicates
Reference
234
Experiment Information Experiment Information Experiment Information Experiment Information Experiment
29.1 ± 3.3 20.9e32 1.48 ± 0.10 0.65e1.3 (8.59 ± 2.61) 103 (11.2 ± 2.2) 103 30.6 ± 3.6 27e36.5 (1.46 ± 0.35) 106 (2.0 ± 0.4) 106 (1.48 ± 0.37) 106 (1.48 ± 0.04) 106
3
This [34] This [34] This [7] This [34] This [7] [9] [29]
235
236
238
236
U U U U U/238U atom ratio
value value value value
Information value
3 3 3 3
3 1
study study study study study
Table 2 236 U/238U atom ratios for the standard reference materials (SRMs). Detail
No. of replicates
236
236
Reference
IAEA-135
Irish Sea sediment
3 1 1 3
11.2 ± 2.2
IAEA-385
Irish Sea sediment
IAEA-447 JSAC 0471
Moss soil Japanese soil
(2.0 ± 0.4) 106 (1.48 ± 0.37) 106 (1.48 ± 0.04) 106 (1.46 ± 0.35) 106 2 108 (2.69 ± 0.71) 108 (1.53 ± 0.07) 107 (2.09 ± 0.73) 108
[7] [9] [29] This [7] This This This
3 4 3
U/238U atom ratio
U (mBq kg1)
SRMs
8.59 ± 2.61 0.1 0.13 ± 0.03 0.65 ± 0.04 (5.18 ± 1.90) 102
study study study study
HNO3 elution (distribution coefficient > 103) [37]. Because trace REEs in seawater have also been studied in the authors' group, the 15 mL elution by 0.1 M HNO3 was selected for further study.
Fig. 1. Effect of the added HF amount during the digestion procedure on the U recovery for IAEA-135 (n ¼ 3).
In addition, AG 1X8 resin or Dowex 1X8 anion exchange resin was also extensively applied for U measurement [5e7,10,12,13,27e29,32]. Recently, published distribution coefficients have indicated that DGA resin has great potential for U purification, since the distribution coefficients (Kd) of U on DGA resin exceed 103 for 6e8 mol L1 HNO3 and are below 10 for low concentration of HNO3 (~0.1 mol L1) [37]. However, application of only the single DGA resin column in the U measurement has been limited [23]. In this study, the elution profiles of U and other matrix elements in DGA resin in different acid matrices were evaluated systematically. As shown in Fig. 2a, 99.65 ± 0.02% recovery of U could be achieved using a 14 mL elution by 0.1 M HNO3. In addition, 99.42 ± 0.02% recovery of U could also be achieved by applying a 16 mL elution by 1 M HCl, as shown in Fig. 2b. Rare earth elements (REEs) are co-eluted with U by 1 M HCl (distribution coefficient z 1), while, REEs remain on the DGA resin during 0.1 M
Fig. 2. Elution profile of U using DGA resin.
Please cite this article in press as: G. Yang, et al., Determination of 236U in environmental samples by single extraction chromatography coupled to triple-quadrupole inductively coupled plasma-mass spectrometry, Analytica Chimica Acta (2016), http://dx.doi.org/10.1016/j.aca.2016.09.033
G. Yang et al. / Analytica Chimica Acta xxx (2016) 1e7
5
The simple chemical separation of U shown in Fig. 4, incorporating HF þ HNO3 þ HClO4 total dissolution and DGA resin single column chromatography, was performed in the following evaluations. 3.3. Optimization of ICP-MS/MS Until now, ICP-MS/MS had been applied for the analysis of the U/238U atom ratio in only one seawater sample [25]. Therefore, analyses of more samples, especially samples with a high matrix effect, are desired to widen the application of ICP-MS/MS in environmental studies and the application of 236U/238U atom ratio as a geochemical tracer. In order to improve the sensitivity of ICP-MS/ MS, the APEX-Q desolvating introduction system is expected to be used for the 236U/238U atom ratio analysis. The signal intensity of U decreased when increasing the O2 gas flow rate, due to the formation of UOþ. Later, 0.1 mL min1 was selected because at this flow rate most of the U reacts with O2 to form UOþ with a stable signal. For 1 ng mL1 of U standard solution, the intensity proportions were estimated as Uþ:UHþ:UOþ:UOHþ ¼ 0.089:0.000:0.910:0.001. In order to check the peak tailing effect of 238U on the analysis of 236 U, 1 ng mL1 of U standard solution was introduced into the ICPMS/MS system. 238UHþ was also monitored to determine the nUHþ interference on the nþ1Uþ analysis. When the Q1 was set for m/z 238, no signal intensities were found at m/z 236 and 239 (UHþ) after Q2 (Fig.S2). Fig. 5 shows the mass spectra of (a) 100 ng mL1 U standard solution (SPEX-XSTC 622) and (b) a typical soil sample with the final introduction concentration of ~1000 ng mL1. Only background noise was observed at m/z 253 for 100 ng mL1 U standard solution. In the present study, a high abundance sensitivity (253X/238U16O) of ~109 was usually obtained for the soil samples. Similar abundance sensitivity was also reported for ICPMS/MS with a different desolvating inlet system (ARIDUS, Cetac Technologies, Omaha, NE, USA) [25]. All these indicated the 236
Fig. 3. Elution profile of Fe and Ca using DGA resin.
For the U analysis by ICP-MS/MS, after applying the total dissolution procedure for soil and sediment samples, the major matrix elements remaining were Si, Fe, Ca, Na, and Mg. No Na and Mg adsorbed on the DGA resin [37]. Furthermore, Si can be removed via reaction with HF after digestion. As shown in Fig. 3, 93.4% of Fe and 46.1% of Ca went through the DGA resin directly during the loading with 6 M HNO3. After rinsing with 20 mL of 6 M HNO3, the Fe concentration could be cut to
Contents lists available at ScienceDirect
Analytica Chimica Acta journal homepage: www.elsevier.com/locate/aca
Determination of 236U in environmental samples by single extraction chromatography coupled to triple-quadrupole inductively coupled plasma-mass spectrometry Guosheng Yang a, b, c, Hirofumi Tazoe a, Masatoshi Yamada a, * a b c
Department of Radiation Chemistry, Institute of Radiation Emergency Medicine, Hirosaki University, 66-1 Hon-cho, Hirosaki, Aomori, 036-8564, Japan Division of Nuclear Technology and Applications, Institute of High Energy Physics, Chinese Academy of Sciences, China Beijing Engineering Research Center of Radiographic Techniques and Equipment, Beijing, 100049, China
h i g h l i g h t s
g r a p h i c a l a b s t r a c t
A simple 236U/238U analytical method has been developed. The separation required just one DGA column chromatography. 236U/238U atom ratios in soil were measured by ICP-MS/MS. 236U/238U atom ratios of (1.50 e13.5) 108 were observed in Japanese samples. activities of (2.25 236U e14.1) 102 mBq kg1 were found in Japanese samples.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 1 August 2016 Received in revised form 8 September 2016 Accepted 23 September 2016 Available online xxx
In order to measure trace 236U and 236U/238U in environmental samples with a high matrix effect, a novel and simple method was developed that makes the digestion and purification procedures compatible with advanced triple-quadrupole inductively coupled plasma-mass spectrometry. A total dissolution of sample with HF þ HNO3 þ HClO4 was followed by chromatographic separation with a single resin column containing normal type DGA resin (N,N,N0 ,N’-tetra-n-octyldiglycolamide) as the extractant system. The analytical accuracy and precision of 236U/238U ratios, measured as 236U16Oþ/238U16Oþ, were examined by using the reference materials IAEA-135, IAEA-385, IAEA-447, and JSAC 0471. The low method detection limit (3.50 106 Bq kg1) makes it possible to perform routine monitoring of environmental 236 U due to global fallout combined with the Fukushima Daiichi Nuclear Power Plant accident fallout (>105 Bq kg1). Finally, the developed method was successfully applied to measure 236U/238U ratios and 236 U activities in soil samples contaminated by the accident. The low 236U/238U atom ratios ((1.50 e13.5) 108) and 236U activities ((2.25e14.1) 102 mBq kg1) indicate 236U contamination was mainly derived from global fallout in the examined samples. © 2016 Elsevier B.V. All rights reserved.
Keywords: 236 U activity 236 U/238U Triple-quadrupole inductively coupled plasma-mass spectrometry Soil Fukushima Daiichi nuclear power plant accident
1. Introduction Naturally and artificially occurring 234U, 235U, and 238U have been extensively studied in previous studies [1e5]. However, the * Corresponding author. E-mail address: [email protected] (M. Yamada).
evidence for U source identification from 235U and 238U is limited, especially in soil samples, because the anthropogenic U has been strongly diluted by the large amount of natural U [3]. 236U has an extremely low natural abundance (236U/238U: ~1014 atoms) in the general environment [4,5]. The contribution from a decay of 240Pu in global fallout is also negligible compared to the directly
http://dx.doi.org/10.1016/j.aca.2016.09.033 0003-2670/© 2016 Elsevier B.V. All rights reserved.
Please cite this article in press as: G. Yang, et al., Determination of 236U in environmental samples by single extraction chromatography coupled to triple-quadrupole inductively coupled plasma-mass spectrometry, Analytica Chimica Acta (2016), http://dx.doi.org/10.1016/j.aca.2016.09.033
2
G. Yang et al. / Analytica Chimica Acta xxx (2016) 1e7
introduced artificial 236U since 240Pu has a long half-life (6564 y) and it is at a low concentration level in the environment. Therefore, presently, the ultra-trace amount of 236U found in environmental and biological samples has been mainly introduced by nuclear weapon explosions (global fallout), nuclear power generation and reprocessing of its fuel, and nuclear reactor accidents. The 236U/238U atom ratio varies with reactor, weapon, and fuel type [6e9]. Thus, 236 238 U/ U is potentially useful as a tracer for the source identification of anthropogenic U present in the environment. The Fukushima Daiichi Nuclear Power Plant (FDNPP) reactors used UO2 and mixed oxide ((U,Pu)O2) fuels. On March 11, 2011, the abundance of 235U in the Unit 1 reactor was estimated as 1.7%, and it was 1.9% for Unit 2 and 3 reactors, which are obviously different from the percent abundance of natural 235U (~0.72%) [2]. However, relatively high concentrations of natural U in Japanese soil (about 1e3 ppm) have prevented identification of the 235U/238U fingerprint from the FDNPP accident, especially for the granitoid areas in Fukushima Prefecture [1,10]. In addition, the 235U/238U atom ratios in plant samples obtained from Fukushima Prefecture after the FDNPP accident also can not reflect the release of U from this accident [2]. 236U concentration can provide additional information on nuclear activities in the contaminated areas, and the 236U/238U atom ratio can provide a far more sensitive tool for the detection of nuclear contamination with respect to other U atom ratios (e.g. 235 238 U/ U) [3,10,11]. Therefore, in this situation, 236U/238U is very appropriate and far more sensitive for demonstrating anthropogenic U contamination in the terrestrial environment from the FDNPP accident [12]. Although Sakaguchi et al. [13] measured 236U in river and sea water samples in the vicinity of the FDNPP, they could not evaluate the source. However, higher 236U/238U atom ratios up to ~106 were observed in airborne samples in the weeks immediately after the FDNPP accident [3]. Furthermore, the isotopic fractions of 236U isotopes in “black substances” collected along roadsides in Fukushima Prefecture revealed the substances contained U fallout from the FDNPP accident [10]. However, the 236U data in the environmental samples illustrating the U release from the plant have been limited until now, especially in soil. Thus, the characteristics of U as a refractory element that was emitted have not been fully clarified in the environment, even 5 years after the accident. The measurement of the trace 236U in the environment presents a challenge to the clarification. For the measurement of 236U, inductively coupled plasma-mass spectrometry (ICP-MS), thermal ionization mass spectrometry (TIMS) and accelerator mass spectrometry (AMS) have usually been employed. AMS is presently the method with the highest detection sensitivity for 236U. However, due to the high instrument price, there are only about 110 AMS facilities worldwide, and most of them are mainly applied to the routine analysis of 14C for dating purposes; only about ten can be used in the study of 236U [3,5,6,9e11,13e19]. TIMS is less preferred than ICP-MS because of greater difficulties in its sample preparation and measurement techniques. To date, ICP-MS is the most commonly applied mass spectrometric method for the investigation of actinides in the environment [20], and it has a detection limit for the 236U/238U atom ratio of about 108 [21]. In addition, until now, the only 236U data for the FDNPP accident were from the Vienna Environmental Research Accelerator (VERA) Laboratory at the University of Vienna, Austria, not from Japanese laboratories. Therefore, it is difficult to perform routine monitoring for 236U due to the FDNPP accident contamination, let alone to establish the 236U/238U database for future applications. However, from the viewpoint of risk assessment and protection against potential nuclear accidents, it is desirable to obtain sufficient 236U/238U background data, since more nuclear power plants and nuclear waste processing facilities
are being or will be built in the foreseeable future in the Northeast Asian region. Therefore, ICP-MS may be an alternative apart from AMS for routine monitoring of 236U. The ICP-MS analysis of 236U in environmental samples, however, remains a great challenge due to the matrix effect, the presence of polyatomic interferences such as 235UHþ, and the tailing effect from 238 U and 235U. Therefore, extraction chromatography, e.g. using UTEVA® and TRU resins, has been applied in tandem for the separation and purification of U from the remaining constituents of the matrix with a smaller solvent volume compared to ionic chromatography [22]. Among available extraction chromatographic resins, normal type DGA resin (N,N,N0 ,N0 -tetra-n-octyldiglycolamide as the extractant system) is another potential resin for U purification, but it requires further studies especially for that condition in which only a single DGA resin column is used [23]. In terms of the detection technique, collision/reaction cell technology was developed to overcome polyatomic interferences like 235UHþ and the tailing effect from 238U and 235U, since it is possible to get effective partitioning of ionemolecule reactions with O2 gas between Uþ and UHþ (Uþ þ O2 / UOþ þ O; UHþ þ O2 / UOþ þ OH) [24,25]. However, to apply collision or reaction cell technology with confidence, more environmental samples should be measured, including those with a high matrix effect. Therefore, an appropriate purification technique by DGA resin compatible with the confirmed collision/reaction cell technology for U analysis may be a potential solution for trace 236U analysis in high matrix effect samples. In the present study, a systematic analysis was made of normal dissolution (acid leaching) and total dissolution (complete digestion) procedures for U atom ratio measurements from environmental samples. Almost no loss of U was achieved during the DGA resin purification procedure. For triple-quadrupole inductively coupled plasma-mass spectrometry (ICP-MS/MS) analysis, an almost interference-free and non-peak tailing spectrum was obtained by measuring UOþ after adding O2 to an octopole collision/ reaction cell located between two quadrupole mass filters. The established method was first examined by measuring four soil and sediment reference materials, and second it was applied to the analysis of soil samples contaminated by the FDNPP accident. 2. Experimental section 2.1. Instrumentation An Agilent 8800 ICP-MS/MS (Agilent Technologies, Santa Clara, CA, USA) featuring an octopole collision/reaction cell situated between two quadrupole mass filters (first, Q1; second, Q2) was employed for analysis of U atom ratios. The introduction system was a high efficiency sample introduction system (APEX-Q, Elemental Scientific, Omaha, NE, USA) equipped with a PFA MicroFlow nebulizer. Flowing nitrogen (10 mL min1) was used to increase transport efficiency and signal stability. The sample uptake rate was ~0.5 mL min1. The ICP-MS/MS was optimized on a daily basis using 1 ng mL1 of SPEX-XSTC 622 standard solution containing U, and the optimized operation parameters are summarized in Table S1. Inductively coupled plasma-optical emission spectrometry (ICP-OES, SPECTRO BLUE TI, SPECTRO Analytical Instruments, Germany) was employed in the U purification procedure to monitor its major interfering elements. 2.2. Reagents and materials Analytical grade HNO3, HF, and HClO4 were obtained from Kanto Chemical Co., Inc. (Tokyo, Japan). Ultrapure grade HNO3 obtained from Tama Chemicals (Tokyo, Japan) and Milli-Q water (>18.2 MU cm1) were used for preparation of the final sample
Please cite this article in press as: G. Yang, et al., Determination of 236U in environmental samples by single extraction chromatography coupled to triple-quadrupole inductively coupled plasma-mass spectrometry, Analytica Chimica Acta (2016), http://dx.doi.org/10.1016/j.aca.2016.09.033
G. Yang et al. / Analytica Chimica Acta xxx (2016) 1e7
solution prior to ICP-MS/MS measurement. The 2 mL pre-packed cartridges containing DGA resin (particle size: 50e100 mm) were purchased from Eichrom Technologies LLC (Lisle, IL, USA). 10 mg L1 U standard stock solution in 5% HNO3 (SPEX-XSTC 622) was obtained from SPEX CertiPrep (Metuchen, NJ, USA). High purity O2 (>99.999%) was obtained from Taiyo Nippon Sanso Corp. (Tokyo, Japan). Marine sediment standard reference material (IAEA-135) was used to investigate the yield of every procedure. Other soil (IAEA-447 and JSAC 0471) and sediment (IAEA-385) standard reference materials were also applied to examine the method via measuring U atom ratios in samples with a high matrix effect. 2.3. U isotopes analysis About 1 g samples were ashed in a muffle oven at 450 C for 2 h to decompose organic matter. First, acid leaching (8 M HNO3 or concentrated HNO3) and total dissolution (HF þ HNO3 þ HClO4) procedures were evaluated as acid attack techniques for U dissolution using IAEA-135. Acid leaching was performed in PFA jars with lids (Savillex, Eden Prairie, MN, USA) on a hot plate at 180 C for 1 d. After filtration, the filtrate was evaporated to dryness. For the total dissolution, the residue was dissolved in 5 mL of 46% HF and 1 mL of 70% HClO4, and evaporated to near dryness again. This step was repeated five times more to remove remaining Si in the sample. After that, the residue was dissolved into 5 mL of 61% HNO3 and evaporated to dryness to remove the residual HF. For the normal dissolution, it is not necessary to perform Si and HF removal. After that, the sample was redissolved into 10 mL of 6 M HNO3 for chromatographic purification using the 2 mL prepacked cartridge containing DGA resin. Using 15 mL of 0.1 M HNO3, U was eluted from the resin, the eluent was evaporated to near dryness, and the residual was dissolved in 1.5 mL of 4% HNO3. For 238U concentration measurement via ICP-MS, a 20 mL aliquot was taken out and diluted with 4% HNO3 by a dilution factor of 2000 in order to get the sample concentrations within the calibration interval (0.01e1 ng mL1). The other fraction was analyzed for 236U/238U, 234 238 U/ U, and 235U/238U atom ratios via ICP-MS/MS. The signal intensities of 238U procedure blanks, incorporated during chemical treatment, were in the range of errors (1s for all data in this study) of the signal intensities of samples. Therefore, the digestion, purification, and dilution procedures did not introduce significant contamination for the U analysis. 3. Results and discussion 3.1. Total dissolution of U It is well known that anthropogenic Pu originating from global fallout can be recovered by acid leaching [20,26]. For soil and sediment samples, various kinds of acid leaching have been applied for 236U measurement previously. Some studies have demonstrated that 8 M HNO3 was sufficient for 236U leaching from soil samples [6,13,27e29]. Moreover, concentrated HNO3 was also used for 236U fallout leaching from soil matrix [5,8,30]. Since there is only a trace amount of 236U in environmental samples and potentially 236U may be present in the yield tracers, e.g. 232U and 233U, total dissolution and complete recovery during purification were desired for the accurate measurement of 236U with no addition of tracer. Furthermore, the aim of this study is 236U/238U atom ratio analysis for source identification studies in the future. To compare 236U/238U atom ratio among the presently studied samples and with other studies, it is better to release all of the U from the soil samples since 238 U also exists in the inner fractions in soil. A mineral acid mixture of HF with HNO3 or HCl was utilized for the total dissolution of 236U from soil and sediment samples [6,7,10,11,31e33]. In addition,
3
HClO4 and H3BO3 were applied synchronously to avoid the formation of insoluble fluoride. In order to get compatibility with the following DGA resin purification and ICP-MS/MS analysis, a systematic evaluation of normal dissolution (8 M or concentrated HNO3) and total dissolution (HF þ HNO3 þ HClO4) procedures for 236 U measurements from environmental samples was performed. For IAEA-135, screening of the results reported for the different U isotopes showed that there is a difference between the results obtained by using either the normal dissolution procedure or the total dissolution procedure [34]. For the normal dissolution procedure, 8 M HNO3 or concentrated HNO3 was considered to be an appropriate reagent for the recovery of global fallout 236U in soil and sediment samples, since this normal dissolution procedure was able to solubilize U held on mineral surfaces and in secondary phases (inner fractions), rather than attempting dissolution of 236U in primary minerals to avoid any natural U dilution [5,6,8,11,12,15,16,28e30]. Until now, the available 236U activities in the reference materials have been mostly deduced from 238U activities and 236U/238U atom ratios [7,9], as shown in Tables 1 and 2. Therefore, the yield of U isotopes during the digestion procedure was compared via 238U first. In the present study, for normal dissolution, concentrated HNO3 was a better acid for complete recovery of U isotopes in IAEA-135, since the measured 12.3 ± 1.1 Bq kg1 of 238U (n ¼ 3) by concentrated HNO3 leaching was in the range of the information values (11.3e18.1 Bq kg1) [34], as shown in Fig.S1. For total dissolution, the HF amount added in the digestion procedure had a significant effect on the recovery of U, as shown in Fig. 1. When 10 mL HF was added, the yield of 238U activity in IAEA-135 increased to 30.6 ± 3.6 Bq kg1 (n ¼ 3), which was in the range of the information values (27.0e36.5 Bq kg1) [34]. On the contrary, incomplete recovery of 238U for 5 mL HF addition (n ¼ 1) may be due to the insufficient acid amount, while the reason for incomplete recovery for more than 10 mL HF addition (n ¼ 1) was not clear. Finally, as shown in Tables 1 and 2, the 236U/238U atom ratio of (1.46 ± 0.35) 106 (n ¼ 3) was observed for total dissolution with 10 mL of 38% HF, 5 mL of 61% HNO3, and 3 mL of 70% HClO4. This value was in agreement with that given by Hotchkis et al. ((1.48 ± 0.37) 106) [9] and Srncik et al. ((1.48 ± 0.04) 106) [29]. However, a somewhat higher value of (2.0 ± 0.4) 106 was obtained by Lee et al. [7] In addition, 8.59 ± 2.61 mBq kg1 (n ¼ 3) of 236U was observed for IAEA-135, which was also somewhat lower than that by Lee et al. (11.2 ± 2.2 mBq kg1) [7]. The higher 236U activities deduced from 236 238 U/ U atom ratios by AMS and 238U activities by other analytical techniques, with different chemical and analytical procedures, may have less accuracy. In the same study [7], a relatively lower 236 238 U/ U atom ratio of (1.8 ± 0.5) 106 (n ¼ 3) and 236U activity of 10.3 ± 3.0 mBq kg1 were obtained by ICP-MS. The wide range of reported results indicates hot particles may be present in IAEA-135, which is believed to have some contamination from the Sellafield reprocessing plant or the 1957 Windscale accident [9]. As shown in Table 1, apart from 236U and 238U, 234U and 235U were also in the range of information values for IAEA-135. In brief, the total dissolution procedure used in the present study was sufficient to release U isotopes from environmental samples. 3.2. Chromatographic purification of U Since ICP-MS/MS and APEX-Q introduction systems are all susceptible to the matrix effect, it is necessary to carry out U purification first for the samples with a high matrix effect. It is well known that U can be separated from the matrix in environmental samples by extraction resins or anion exchange resins. In previous studies, TRU and UTEVA® resins were successfully applied for the separation and purification of U [4,5,7,9,11,16,27e30,33,35,36].
Please cite this article in press as: G. Yang, et al., Determination of 236U in environmental samples by single extraction chromatography coupled to triple-quadrupole inductively coupled plasma-mass spectrometry, Analytica Chimica Acta (2016), http://dx.doi.org/10.1016/j.aca.2016.09.033
4
G. Yang et al. / Analytica Chimica Acta xxx (2016) 1e7
Table 1 U isotopic information obtained for IAEA-135 after applying the total dissolution procedure. IAEA-135
Note
Value (Bq kg1)
No. of replicates
Reference
234
Experiment Information Experiment Information Experiment Information Experiment Information Experiment
29.1 ± 3.3 20.9e32 1.48 ± 0.10 0.65e1.3 (8.59 ± 2.61) 103 (11.2 ± 2.2) 103 30.6 ± 3.6 27e36.5 (1.46 ± 0.35) 106 (2.0 ± 0.4) 106 (1.48 ± 0.37) 106 (1.48 ± 0.04) 106
3
This [34] This [34] This [7] This [34] This [7] [9] [29]
235
236
238
236
U U U U U/238U atom ratio
value value value value
Information value
3 3 3 3
3 1
study study study study study
Table 2 236 U/238U atom ratios for the standard reference materials (SRMs). Detail
No. of replicates
236
236
Reference
IAEA-135
Irish Sea sediment
3 1 1 3
11.2 ± 2.2
IAEA-385
Irish Sea sediment
IAEA-447 JSAC 0471
Moss soil Japanese soil
(2.0 ± 0.4) 106 (1.48 ± 0.37) 106 (1.48 ± 0.04) 106 (1.46 ± 0.35) 106 2 108 (2.69 ± 0.71) 108 (1.53 ± 0.07) 107 (2.09 ± 0.73) 108
[7] [9] [29] This [7] This This This
3 4 3
U/238U atom ratio
U (mBq kg1)
SRMs
8.59 ± 2.61 0.1 0.13 ± 0.03 0.65 ± 0.04 (5.18 ± 1.90) 102
study study study study
HNO3 elution (distribution coefficient > 103) [37]. Because trace REEs in seawater have also been studied in the authors' group, the 15 mL elution by 0.1 M HNO3 was selected for further study.
Fig. 1. Effect of the added HF amount during the digestion procedure on the U recovery for IAEA-135 (n ¼ 3).
In addition, AG 1X8 resin or Dowex 1X8 anion exchange resin was also extensively applied for U measurement [5e7,10,12,13,27e29,32]. Recently, published distribution coefficients have indicated that DGA resin has great potential for U purification, since the distribution coefficients (Kd) of U on DGA resin exceed 103 for 6e8 mol L1 HNO3 and are below 10 for low concentration of HNO3 (~0.1 mol L1) [37]. However, application of only the single DGA resin column in the U measurement has been limited [23]. In this study, the elution profiles of U and other matrix elements in DGA resin in different acid matrices were evaluated systematically. As shown in Fig. 2a, 99.65 ± 0.02% recovery of U could be achieved using a 14 mL elution by 0.1 M HNO3. In addition, 99.42 ± 0.02% recovery of U could also be achieved by applying a 16 mL elution by 1 M HCl, as shown in Fig. 2b. Rare earth elements (REEs) are co-eluted with U by 1 M HCl (distribution coefficient z 1), while, REEs remain on the DGA resin during 0.1 M
Fig. 2. Elution profile of U using DGA resin.
Please cite this article in press as: G. Yang, et al., Determination of 236U in environmental samples by single extraction chromatography coupled to triple-quadrupole inductively coupled plasma-mass spectrometry, Analytica Chimica Acta (2016), http://dx.doi.org/10.1016/j.aca.2016.09.033
G. Yang et al. / Analytica Chimica Acta xxx (2016) 1e7
5
The simple chemical separation of U shown in Fig. 4, incorporating HF þ HNO3 þ HClO4 total dissolution and DGA resin single column chromatography, was performed in the following evaluations. 3.3. Optimization of ICP-MS/MS Until now, ICP-MS/MS had been applied for the analysis of the U/238U atom ratio in only one seawater sample [25]. Therefore, analyses of more samples, especially samples with a high matrix effect, are desired to widen the application of ICP-MS/MS in environmental studies and the application of 236U/238U atom ratio as a geochemical tracer. In order to improve the sensitivity of ICP-MS/ MS, the APEX-Q desolvating introduction system is expected to be used for the 236U/238U atom ratio analysis. The signal intensity of U decreased when increasing the O2 gas flow rate, due to the formation of UOþ. Later, 0.1 mL min1 was selected because at this flow rate most of the U reacts with O2 to form UOþ with a stable signal. For 1 ng mL1 of U standard solution, the intensity proportions were estimated as Uþ:UHþ:UOþ:UOHþ ¼ 0.089:0.000:0.910:0.001. In order to check the peak tailing effect of 238U on the analysis of 236 U, 1 ng mL1 of U standard solution was introduced into the ICPMS/MS system. 238UHþ was also monitored to determine the nUHþ interference on the nþ1Uþ analysis. When the Q1 was set for m/z 238, no signal intensities were found at m/z 236 and 239 (UHþ) after Q2 (Fig.S2). Fig. 5 shows the mass spectra of (a) 100 ng mL1 U standard solution (SPEX-XSTC 622) and (b) a typical soil sample with the final introduction concentration of ~1000 ng mL1. Only background noise was observed at m/z 253 for 100 ng mL1 U standard solution. In the present study, a high abundance sensitivity (253X/238U16O) of ~109 was usually obtained for the soil samples. Similar abundance sensitivity was also reported for ICPMS/MS with a different desolvating inlet system (ARIDUS, Cetac Technologies, Omaha, NE, USA) [25]. All these indicated the 236
Fig. 3. Elution profile of Fe and Ca using DGA resin.
For the U analysis by ICP-MS/MS, after applying the total dissolution procedure for soil and sediment samples, the major matrix elements remaining were Si, Fe, Ca, Na, and Mg. No Na and Mg adsorbed on the DGA resin [37]. Furthermore, Si can be removed via reaction with HF after digestion. As shown in Fig. 3, 93.4% of Fe and 46.1% of Ca went through the DGA resin directly during the loading with 6 M HNO3. After rinsing with 20 mL of 6 M HNO3, the Fe concentration could be cut to
