- Email: [email protected]
Simultaneous determination of the quantity and isotopic ratios of uranium in individual micro-particles by isotope dilution thermal ionization mass spectrometry (ID-TIMS)
Talanta 160 (2016) 600–606
Contents lists available at ScienceDirect
Talanta journal homepage: www.elsevier.com/locate/talanta
Simultaneous determination of the quantity and isotopic ratios of uranium in individual micro-particles by isotope dilution thermal ionization mass spectrometry (ID-TIMS) Jong-Ho Park n, Eun-Ju Choi Nuclear Chemistry Research Division, Korea Atomic Energy Research Institute, 111 Daedeok-daero-989, Yuseong-gu, Daejeon, 34057 Republic of Korea
art ic l e i nf o
a b s t r a c t
Article history: Received 14 July 2016 Received in revised form 27 July 2016 Accepted 1 August 2016 Available online 2 August 2016
A method to determine the quantity and isotopic ratios of uranium in individual micro-particles simultaneously by isotope dilution thermal ionization mass spectrometry (ID-TIMS) has been developed. This method consists of sequential sample and spike loading, ID-TIMS for isotopic measurement, and application of a series of mathematical procedures to remove the contribution of uranium in the spike. The homogeneity of evaporation and ionization of uranium content was confirmed by the consistent ratio of n(233U)/n(238U) determined by TIMS measurements. Verification of the method was performed using U030 solution droplets and U030 particles. Good agreements of resulting uranium quantity, n (235U)/n(238U), and n(236U)/n(238U) with the estimated or certified values showed the validity of this newly developed method for particle analysis when simultaneous determination of the quantity and isotopic ratios of uranium is required. & 2016 Elsevier B.V. All rights reserved.
Keywords: Uranium Thermal Ionization Mass Spectrometry Isotopic Ratio IDMS Particle Analysis
1. Introduction Reliable determination of the quantity and isotopic ratios of nuclear materials, such as uranium (U) and plutonium (Pu), in environmental samples plays an important role in monitoring undeclared nuclear activities for nuclear safeguards and in tracking the origin of samples for nuclear forensics [1–3]. Information on the amounts of U and/or Pu in a sample can be suggestive of deviation from declared activities, while the nuclear history of samples and the corresponding facilities can be tracked using the isotopic ratios. Therefore, the acquisition of both types of information is important for making accurate decisions relating nuclear safeguards. Nuclear material analysis can be classified into two main categories: bulk analysis and particle analysis. The objective of the former is to determine both the amounts and the isotopic ratios of nuclear materials in samples [4–6]. Because all of the nuclear particles in a sample are chemically treated and dissolved in a solution prior to isotopic measurement by inductively coupled plasma mass spectrometry (ICP-MS) or thermal ionization mass spectrometry (TIMS), the isotopic ratios acquired by bulk analysis are just average values. Therefore, the information obtained by bulk analysis provides only a rough clue to the occurrence of n
Corresponding author. E-mail address: [email protected] (J.-H. Park).
http://dx.doi.org/10.1016/j.talanta.2016.08.006 0039-9140/& 2016 Elsevier B.V. All rights reserved.
undeclared nuclear activities. The isotopic ratios of U and/or Pu in individual micro-particles contained in a sample can be determined by TIMS, which provides essential information with which to track the nuclear history of a facility. TIMS combined with the fission track (FT) technique or energy-dispersive spectroscopy in a scanning electron microscope (SEM-EDS) for particle identification is one of the best options to measure the isotopic ratios with high accuracy, precision, and sensitivity [7–11]. Secondary ion mass spectrometry (SIMS) is another method for particle analysis with a fast particle search process and reliable isotopic analysis [12–14]. However, the amount of nuclear materials is simply estimated by measuring the particle size, with the assumption that the micro-particles consist purely of U or Pu and are spherical in shape. When particles deviate from these assumptions, there is a similar deviation of the determined quantity from the true value. Recently, ICP-MS has been utilized for isotopic ratio analysis of individual sub-micrometer plutonium particles by a combination of chemical dissolution and separation processes [15]. Quantitative analysis of Pu is possible by measuring the signal intensity and by utilizing the standard curve. This method is useful when sufficient instrumental sensitivity is ensured and a negligible background from the chemical process is guaranteed. However, quantitative and isotopic ratio analyses by this method are hardly available for individual micro-particles of uranium, because a considerable amount of U background is expected during the chemical process. Isotope dilution mass spectrometry (IDMS) is the most accurate
J.-H. Park, E.-J. Choi / Talanta 160 (2016) 600–606
technique for quantitative analysis [16–18]. In this approach, a known amount of spike reference solution is mixed with a sample solution followed by isotopic analysis using ICP-MS or TIMS. This technique is limited, however, in cases when a sample can hardly be dissolved [17] due to inhomogeneous mixing with the spike. Determination of the quantity of nuclear micro-particles is included in such cases. Furthermore, mixing a sample with the spike results in distortion of the isotopic ratios of the target element. The sample solution is normally divided into two portions: one for mixing with the spike reference solution and the other for isotopic ratio determination. However, if a sample consists of micro-particles, this process is not possible, which means that an additional technique to remove the distortion of the isotopic ratio is required. Kraiem et al. used isotope dilution thermal ionization mass spectrometry (ID-TIMS) to characterize uranium particles, which were monodisperse uranium oxide particles of a known size (ca. 1 mm), from a certified uranium standard solution [17]. The applicability of the aforementioned technique for quantitative analysis of uranium particles was successfully verified, and the determination of uranium isotopic ratios was also performed. However, the quantitative and isotopic analyses were carried out using monodisperse duplicated particles. If the isotopic ratios of U and/ or Pu in individual particles are expected not to be identical, as in the case of typical environmental samples, only one of quantitative and isotopic analyses can be performed, which means that a new method to perform both is required. This paper presents a method to determine both quantity and isotopic ratios of uranium contained in individual micro-particles by ID-TIMS. Homogeneous evaporation and ionization of the uranium from a sample and a spike are described for verification of the application of IDMS to the quantification of uranium in microparticles. A mathematical model and a reasonable assumption were applied to remove the contribution of the spike to the isotopic ratios of uranium in individual micro-particles. The quantitative and isotopic analytical results are presented to validate the newly developed method for particle analysis.
2. Experimental 2.1. Spike and sample materials IRMM-040a (EU-JRC-IRMM), as the certified spike isotopic reference material, was diluted with nitric acid (2 M) to make a spike solution. The amount of 233U in the spike solution was 2.065 2 [ 70.003 6] 10 10 mol g 1. (The numbers in parentheses in this paper indicate expanded uncertainty, Ue ¼ k uc, where k and uc are the coverage factor and the combined standard uncertainty, respectively.). U030 (National Bureau of Standards, USA), a certified reference material, was used for the test samples to validate the test method. The test samples were divided into two main types: nitric acid (2 M) solution samples of U030 (called ‘U030 solution’) and microparticle samples of U030 powder (highly purified oxide, U3O8). The amount of 238U in the U030 solution was determined preliminarily to be 2.261 [7 0.014] 10 10 mol g 1 by IDMS. 2.2. Sample-spike mixture preparation Sample loading and spike addition were sequentially performed on a zone-refined rhenium filament (Thermo Fisher Scientific, Bremen, Germany). Firstly, 1 μL of U030 solution was transferred onto a filament using a micro-pipette; the relative uncertainty of this volume was 1.5%. The calculated amount of 238U in the transferred U030 solution was 57.2 [ 71.8] pg. Once the transferred U030 solution was completely dried by heat created by
601
a current of 0.6 A, 1 μL of the spike solution was transferred onto the same spot where the dried U030 solution was located, and was then dried. Our preliminary experiment showed that mixing by the aforementioned process resulted in inhomogeneous evaporation and ionization of 233U and 238U, and the signal intensity ratio of n(233U)/n(238U) was not constant throughout a measurement. To resolve this problem, re-dissolution of the mixture was performed by adding 1 M HNO3 on the spot where the mixture was located, followed by drying. Two volumes (i.e., 2 μL) of nitric acid were added to cover the spot sufficiently for complete re-dissolution of the mixture. Re-dissolution was repeated five times, and then the mixture was fixed on the filament by applying a current of 1.8 A for 30 s; five samples were prepared (S1–S5). In the case of micro-particle samples, a micro-manipulator system equipped in a scanning electron microscope (SEM) was utilized to transfer a single particle (ca. 2 mm) to a filament (Fig. 1). The rest of the procedure was the same as for the case of the U030 solution; five samples were prepared (P1–P5). 2.3. Isotopic measurement and data analysis The isotopic measurement using TIMS (Triton Plus; Thermo Fisher Scientific) adopting the continuous heating method was performed as described elsewhere [6,11,18–20]. A multi-ion counter system, consisting of three secondary electron multipliers (SEMs) and two compact discrete dynode detectors (CDDs), was used for simultaneous measurement. The detector configuration used in this study is described in Table 1. Detection efficiencies for the five ion counters were adjusted using the ion signals of 187Re (ca. 300,000 cps). The mass bias for the isotopes, with the exception of 233U, was corrected by isotopic measurement of a reference material (U200; National Bureau of Standards, USA). The mass bias for 233U was not corrected due to the lack of availability of an appropriate reference material. Only the data sets, whose 238U þ intensity fell in the range of 5– 100% with respect to the maximum 238U þ intensity, were considered as the valid data sets to avoid distortion of the resulting isotopic ratios and unnecessarily large uncertainty arising from relatively small signal intensity. The weights for the individual data sets were considered to calculate the weighted means and the weighted standard errors of the isotopic measurements. The uncertainty was estimated according to GUM (Guide to the Expression of Uncertainty in Measurement), in compliance with ISO/IEC Guide 98-3 [21,22]. All of the analytical processes were performed in a clean facility, which was controlled to ISO 5 (class 100) and ISO 6 (class 1000) levels to avoid sample contamination.
3. Results and discussion 3.1. Homogeneous evaporation and ionization of uranium The mass spectrometric behavior, as well as the chemical behavior of 233U from the spike and 238U from a sample, must be identical for valid application of IDMS. Figs. 2(a) and 3(b) are the examples of 233U þ and 238U þ intensity profiles during TIMS measurements for a mixture of a U030 solution droplet and the spike, and a mixture of a U030 particle and the spike, respectively. The dash-dotted lines indicate the 5% levels of the maximum intensities of 238U þ in each profiles, and the valid data sets are above them. The profiles shows that the evaporation and ionization of the two isotopes are identical during a measurement. The change in the isotopic ratios of n(233U)/n(238U) in the range of valid data sets, as shown in Figs. 2(b) and 3(b), were approximately 5–8%. This amount of change is typical when the mass fractionation of
602
J.-H. Park, E.-J. Choi / Talanta 160 (2016) 600–606
Fig. 1. Micro-particle transfer to a filament using a micromanipulation system.
TIMS is considered, and is not responsible for the different mass spectrometric behavior of 233U and 238U. Therefore, although homogeneous mixing of the spike and a sample was not guaranteed, no inhomogeneity in evaporation and ionization of the two isotopes was observed.
Table1 Detector configuration of TIMS in this study. Isotope
234
IC4 CDD
IC3 SEM
U
Name Type
Ion Signal Intensity (cps)
Detector
233
U
235
236
238
IC2 SEM
IC1 SEM
IC5 CDD
U
U
U
800.0k
(a) 600.0k U U
400.0k 200.0k
5%
0.0 0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
1.00
n(
233
U)/n(
238
U)
0.95
(b)
0.90 0.85 0.80 0.75 0.70 0.65 2600
2800
3000
3200
3400
3600
3800
4000
Current Applied to Evaporational Filament (mA) Fig. 2. The profiles of 233U þ and 238U þ intensities during TIMS measurement for a mixture of a U030 solution droplet and the spike (a) and the change in n(233U)/n(238U) in the range of valid data sets (b). The dash-dotted line in (a) indicates the 5% level of the maximum intensity of 238U þ to determine the range of valid data sets.
J.-H. Park, E.-J. Choi / Talanta 160 (2016) 600–606
603
Fig. 3. The profiles of 233U þ and 238U þ intensities during TIMS measurement for a mixture of a U030 particle and the spike (a) and the change in n(233U)/n(238U) in the range of valid data sets (b). The dash-dotted line in (a) indicates the 5% level of the maximum intensity of 238U þ to determine the range of valid data sets.
3.2. Calculation of quantity and isotopic ratios The amount of 238U in a sample is determined by the following equations [15–17]:
(
csample
238
)
U =
rspike − rmixture 1 + rsample Vspike⋅dspike ⋅ ⋅ rmixture − rsample 1 + rspike Vsample⋅dsample
(
⋅cspike
233
U
)
(
238
(
238
csample
(1)
or
asample
can be completely ruled out, for example, by periodic monitoring of the spike solution in compliance with a quality assurance and quality control program. Under this assumption, Eqs. (1) and (2) are reduced as follows, respectively:
(
)
(
233
)
U ,
(2)
where csample(238U) and cspike(233U) are the amounts content (mol g 1) of 238U in a sample and 233U in the spike solution; asample(238U) and aspike(233U) are the amounts (mol) of 238U in a sample and 233U in the spike solution added to a sample, respectively; rsample, rspike, and rmixture are the amount ratio n(233U)/n (238U) in a sample, the spike, and a mixture, respectively; Vsample and Vspike are the volumes of a sample and the spike, respectively; and dsample and dspike are the densities of a sample and the spike, respectively. The amounts of 238U and 233U [i.e., asample(238U) and aspike(233U), respectively] are the products of the corresponding volume (Vsample and Vspike), density (dsample and dspike), and the amount contents [csample(238U) and cspike(233U)], respectively, and contain their propagated uncertainties. To acquire the three isotopic ratios (rsample, rspike, and rmixture), the uranium isotopic measurement by TIMS must be performed for a sample, the spike, and a mixture in normal cases. However, the measurement for a sample to obtain rsample is not possible in this case as no replica is available. Therefore, the following assumption is necessary for application of ID-TIMS to the proposed method: no 233U nuclide is contained in a sample, which enables us to set rsample to zero. Because 233U is a fissile isotope of uranium produced by the neutron irradiation of 232Th (thorium-232), its natural abundance is negligible. Therefore, this assumption is reasonable. In addition, for practical purposes, the certified value of the spike material can be used for the amount ratio n(233U)/n (238U) in the spike (i.e., rspike) if the possibility of contamination
rspike − rmixture rmixture
⋅
Vspike⋅dspike 1 ⋅ ⋅cspike 1 + rspike Vsample⋅dsample
(
233
U
)
(3)
or
asample rspike − rmixture 1 + rsample 238 U = ⋅ ⋅aspike rmixture − rsample 1 + rspike
)
U =
)
U =
rspike − rmixture rmixture
⋅
1 ⋅aspike 1 + rspike
(
233
)
U .
(4)
The isotopic ratios of n(234U)/n(238U), n(235U)/n(238U), and n (236U)/n(238U) in a mixture (Rmixture) are the combination of the corresponding ratios of uranium in a sample (Rsample) and the spike (Rspike) with a contribution factor (s). Therefore, Rsample was determined as follows:
Rsample =
R mixture − (1 − s )⋅Rspike (5)
s
where
s=
asample(238U ) a mixture(
238
U)
aspike(238U )
=1−
a mixture(238U )
The total amount of culated using Eq. (7).
.
238
(6) 238
U in a mixture [amixture(
U)] was cal-
a mixture(238U ) = asample(238U ) + aspike(238U ) = asample(238U ) + aspike(233U )⋅
1 rspike
(7)
The amount and the amount content of total uranium in a sample [asample(U) and csample(U), respectively] are determined based on the amount and the amount content of 238U, and the isotope ratios calculated above as follows:
asample(U ) = asample
(
238
)(
U ⋅ 1+
(
csample(U ) = csample(238U )⋅ 1 +
∑ Rsample),
∑ Rsample).
(8) (9)
604
J.-H. Park, E.-J. Choi / Talanta 160 (2016) 600–606
Table 2 The amount of total U and the isotopic ratios of U in the U030 solution droplets.
Estimated Value Certified Value S1 S2 S3 S4 S5 Mean value Relative Unc.a (%) Accuracyb (%) RSDc (%)
Total U Isotopic Ratio amount (pg) n (234U)/n (238U) ( 10 4)
n (235U)/n (238U) ( 10 2)
n (236U)/n (238U) ( 10 4)
57.2 [ 7 1.8]
–
–
–
–
1.961 [7 0.010] 1.84 [ 7 0.57] 1.55 [ 7 0.91] 1.77 [ 70.42] 1.86 [ 7 0.45] 2.12 [ 7 0.45] 1.83 31
3.143 0 [7 0.003 0] 3.16 [ 70.21] 3.13 [ 70.21] 3.13 [ 70.21] 3.13 [ 70.21] 3.12 [ 7 0.21] 3.13 6.7
2.105 [ 70.010] 2.21 [ 7 0.18] 2.20 [ 7 0.16] 2.21 [ 7 0.15] 2.21 [ 7 0.15] 2.20 [ 7 0.15] 2.21 7.2
6.8 11
0.34 0.50
9.4 0.33
53.5 54.4 60.6 56.8 56.5 56.4 3.0
[7 1.7] [7 1.6] [7 1.9] [7 1.7] [7 1.7]
1.5 4.9
The numbers in parentheses indicate expanded uncertainty, Ue ¼ k uc , where uc is the combined uncertainty and k ¼ 2 with 95% confidential level. a Relative Unc. is the relative uncertainty defined as (Uncertainty)/(Mean Value)*100. b Accuracy is defined as (Measure value – Certified value)/(Certified value) *100. c RSD is the relative standard deviation.
3.3. Determination of quantity and isotopic ratios of U in U030 solution samples Table 2 and Fig. 4 show the amount of total U and the isotopic ratios of U contained in the individual U030 solution droplets (ca. 64 62
1 μL) determined by the proposed method. The estimated amount of transferred uranium on a filament (estimated value) and the certified isotopic ratios (certified value) are also shown. The uncertainty ranges of the U quantity determined by IDTIMS overlapped with the estimated value. The accuracy ranged from 6.5% to 6.0%, the average of which was 1.5%. The relative standard deviation (RSD) of the resulting amounts was 4.9% and only one measurement (S1) deviated slightly from the uncertainty range of the estimated value. Although the performance of the quantification by the proposed method was not as accurate and precise as that by normal IDMS, it was still acceptable for determination of the amount of U, considering that the amount is at an ultra-trace level (ca. 50 pg). For example, the measurement goal of the relative expanded uncertainty of U total mass for the analysis of bulk environmental samples needs to be less than 40% [23]. The isotopic ratios of n(235U)/n(238U) and n(236U)/n(238U) were fairly consistent with the corresponding certified values within the uncertainty ranges. The deviation of n(236U)/n(238U) was considered to be mostly due to the peak tailing contribution of 238U to 236 U, because positive bias was consistently observed [24–26]. The RSDs of n(235U)/n(238U) and n(236U)/n(238U) were 0.50% and 0.30%, respectively, indicating high repeatability. The isotopic ratio of n (234U)/n(238U) were also consistent with the certified value. However, the determination was not highly reliable since the relative uncertainty and RSD were somewhat high (31% and 11%, respectively). This is because the spike material used in this study contained a considerable amount of 234U (ca. 0.93% of 233U), which blocks the contribution of 234U from a sample when mixed for IDTIMS. As shown in Eq. (5), Rmixture is determined mainly by Rspike and the contribution of Rsample becomes minor when Rspike is considerably greater than Rsample. Determination of n(234U)/n(238U) is included in such case because Rspike is 1.137 287 0.000 24 and -4
2.8x10
(a)
-4
2.6x10
(b)
-4
60
2.4x10 Upper Range
-4
2.2x10
58
-4
2.0x10
-4
1.8x10
56 Lower Range
-4
1.6x10 54
-4
1.4x10
-4
52
U Qauntity (pg)
-2
-4
3.0x10
(c)
-4
234
238
236
238
n( U)/n( U)
-4
1.0x10
50 4.5x10
1.2x10
(d)
2.8x10
-2
4.0x10
-4
2.6x10
-4
2.4x10
-2
3.5x10
-4
2.2x10 -2
3.0x10
-4
2.0x10
-4
1.8x10
-2
2.5x10
-4
1.6x10 -2
235
2.0x10
238
n( U)/n( U) S1
S2
S3
S4
S5
-4
1.4x10
n( U)/n( U)
-4
1.2x10
S1
S2
S3
S4
S5
Fig. 4. The amount of total U (a) and the isotopic ratios of n(234U)/n(238U) (b), n(235U)/n(238U) (c), n(236U)/n(238U) (d) contained in the U030 solution samples. The dashed line and the dash-dotted lines in (a) are the estimated value of U quantity and the upper/lower ranges of uncertainty of the estimated value, respectively. The dashed lines in (b), (c), and (d) are the certified values.
J.-H. Park, E.-J. Choi / Talanta 160 (2016) 600–606
Table 3 The amount of total U and the isotopic ratios of U in the U030 particles. Total U amount (pg)
P1 P2 P3 P4 P5 Mean value Relative Unc. (%) Accuracy (%) RSD (%)
Isotopic Ratio n (234U)/n (238U) ( 10 4)
n (235U)/n (238U) ( 10 2)
36.5 [7 1.1] 21.0 [ 7 0.9] 25.5 [7 1.1] 21.4 [ 7 0.8] 22.0 [7 1.1] – –
7.1 [ 7 1.2] 5.8 [ 7 7.9] 4.3 [ 7 6.1] 7.9 [7 5.0] 2.5 [ 7 8.9] 5.5 110
3.18 3.14 3.13 3.11 3.15 3.14 8.3
– –
181.4 39.5
0.08 0.81
n (236U)/n (238U) ( 10 4)
[ 7 0.21] [ 7 0.28] [ 7 0.27] [7 0.23] [ 7 0.31]
2.30 2.25 2.35 2.24 2.36 2.30 13
[ 70.18] [ 70.34] [7 0.28] [7 0.27] [ 70.37]
9.4 2.43
The numbers in parentheses indicate expanded uncertainty, Ue ¼ k uc , where uc is the combined uncertainty and k ¼ 2 with 95% confidential level.
Rsample is in the order of 10 4. The accurate value of n(235U)/n(238U) reconfirmed that the U quantity values determined by this method were acceptable because the isotopic ratios are closely related to the quantity information [see Eqs. (5)–(9)]. 3.4. Determination of quantity and isotopic ratios of U in U030 particles The method for the simultaneous determination of the quantity and isotopic ratios of uranium was applied to the particle samples (P1 to P5), the results of which are summarized in Table 3. Because the U030 particles in this study were not well shaped, the accurate estimation of the uranium amount by size measurement was not possible. Instead, roughly 30 pg of uranium (with a high level of uncertainty) was estimated to be contained in individual particles. The amounts of uranium in the U030 particles measured by the proposed method fell into a range close to the rough estimate. There is no direct way to verify whether the measurement results were consistent with real values, as in the case of U030 solution samples. However, indirect verification is possible by checking the isotopic ratios because the quantity information is related to these ratios. Agreement of the isotopic ratios of n(235U)/n(238U) and n(236U)/ n(238U) with the certified values within the uncertainty ranges was observed, as more clearly shown in Fig. 5. The accuracies of individual measurements for n(235U)/n(238U) and n(236U)/n(238U) ranged from 0.5% to 1.1% and from 6.4% to 12.0%, respectively. Although consistent positive bias was observed for n(236U)/n 4.5x10
-2
4.2x10
-2
3.9x10
-2
3.6x10
-2
3.3x10
-2
3.0x10
-2
2.7x10
-2
2.4x10
-2
2.1x10
-2
1.8x10
-2
235
(a)
238
n( U)/n( U)
P1
P2 235
Fig. 5. The isotopic ratios of n(
U)/n(
P3
238
P4 236
U) (a) and n(
U)/n(
238
P5
605
(238U), this was mostly due to the peak tailing effect as in the case of U030 solution samples, and is typically observed in TIMS measurements unless a correction method is applied or a retarding potential quadrupole (RPQ) system is utilized [25]. The relative uncertainty and RSD were 8.3% and 0.81%, respectively, for n (235U)/n(238U), and 13% and 2.43% for n(236U)/n(238U). On the other hand, the isotopic ratio of n(234U)/n(238U) deviated markedly from the certified values. The average accuracy, relative uncertainty, and RSD were 181.4%, 110%, and 39.5%, respectively. As explained above, the large deviation and low precision are caused by the high content of 234U in the spike, which blocked the 234U in the samples. Furthermore, this effect was relatively enhanced for the particle analysis compared with that for the U030 solution droplet analysis, because the amounts of 234U in individual particles were approximately half those in U030 solution droplet samples, while those in the spike were the same. However, selecting a more appropriate spike material, which contains an extremely small amount of 234U, would resolve this problem. Despite the inaccurate results for n(234U)/n(238U) due to the adoption of an inappropriate spike material, the determination of n(235U)/n(238U) and n(236U)/n(238U) using the proposed method is considered to be highly reliable. The analytical result of n(235U)/n (238U) satisfied the accuracy requirements for analysis of particle environmental samples relating to International Atomic Energy Agency (IAEA) safeguards (i.e., less than 1%) [23]. In addition, the accuracy of the isotopic ratio determination is strong evidence that the determination of U quantity was also reliable, as mentioned above. Based on the verification of the analytical performance, the newly developed method can be used for the analysis of uranium in individual particles with an acceptable level of measurement performance. This is especially useful when the quantity and isotopic ratios of uranium must be determined simultaneously and replicated particles are not available.
4. Conclusion A newly developed method to determine the quantity and isotopic ratios of uranium contained in individual micro-particles simultaneously by ID-TIMS is described here. While the determinations of the quantity and isotopic ratios are performed separately by IDMS and TIMS (or ICP-MS), respectively, in normal cases, particle analysis allows only one measurement when replicated particles are not available. Two types of sample (U030 solution droplets and particles) were used for verification of the method. A mixture for ID-TIMS was prepared by sequential loading of a sample and the spike on a TIMS filament, followed by repeated 3.0x10
-4
2.8x10
-4
2.6x10
-4
2.4x10
-4
2.2x10
-4
2.0x10
-4
1.8x10
-4
1.6x10
-4
1.4x10
-4
1.2x10
-4
(b)
P1
236
238
n( U)/n( U)
P2
P3
P4
P5
U) (b) contained in the U030 particles. The dashed lines in (b), (c), and (d) are the certified values.
606
J.-H. Park, E.-J. Choi / Talanta 160 (2016) 600–606
dissolution with nitric acid. Homogeneous evaporation and ionization of 233U and 238U contents were confirmed for validation of the application of ID-TIMS for this purpose. An additional series of mathematical processes was applied to remove the contribution of the spike to the isotopic ratios of uranium in a sample. Fair agreement with the estimated or certified values was observed for the quantity and the isotopic ratios of n(235U)/n(238U) and n(236U)/ n(238U) for the U030 solution droplet sample measurement, the accuracies of which were 1.5%, 0.34%, and 9.4%, respectively, while the reliability of n(234U)/n(238U) analysis was not ensured due to the large amount of 234U in the spike material. The application of this method to particle analysis was verified by the analytical results of the isotopic ratios of n(235U)/n(238U) and n (236U)/n(238U). The deviation and the large RSD in the n(234U)/n (238U) analysis due to the factor mentioned above can be resolved by selecting more appropriate spike materials that contain a lower level of 234U. The analytical method proposed here should be useful for micro-particle analysis when both the quantity and the isotopic ratio of uranium are required, but when no replicated particles are available.
[9]
[10]
[11] [12]
[13] [14]
[15]
[16]
[17]
Acknowledgments This work was supported by the Ministry of Science, ICT, and Future Planning (MSIP) and the Nuclear Safety and Security Commission (NSSC) of South Korea (No. 1405020).
[18]
[19]
References [20] [1] D.L. Donohue, Strengthening IAEA safeguards through environmental sampling and analysis, J. Alloy Compd. 271–273 (1998) 11–18. [2] D.L. Donohue, Strengthended nuclear safeguards, Anal. Chem. 74 (2002) 28A–35A. [3] K. Mayer, M. Wallenius, Z. Varga, Nuclear forensic science: correlating measurable material parameters to the history of nuclear material, Chem. Rev. 113 (2013) 884–900. [4] M. Magara, Y. Hanzawa, F. Esaka, Y. Miyamoto, K. Yasuda, K. Watanabe, S. Usuda, H. Nishimura, T. Adachi, Development of analytical technique for ultra trace amounts of nuclear materials in environmental samples using ICPMS for safeguards, Appl. Radiat. Isot. 53 (2000) 87–90. [5] O.T. Farmer III, K.B. Olsen, M.L. Thomas, S.J. Garofoli, Analysis of IAEA environmental samples for plutonium and uranium by ICP/MS in support of international safeguards, J. Radioanal. Nucl. Chem. 276 (2008) 489–492. [6] J.-H. Park, S. Lee, Y.-G. Ha, S.A. Lee, K. Jeong, M. Lee, K. Song, Bulk analysis of a simulated environmental sample in natural abundance performed in KAERI for nuclear safeguards, J. Radioanal. Nucl. Chem. 303 (2015) 1297–1300. [7] K.G. Heumann, S. Eisenhut, S. Gallus, E.H. Hebeda, R. Nusko, A. Vengosh, T. Walczyk, Recent developments in thermal ionization mass spectrometric techniques for isotope analysis, Analyst 120 (1995) 1291–1299. [8] O. Stetzer, M. Betti, J. Geel, N. Erdmann, J.-V. Kratz, R. Schenkel, N. Trautmann,
[21] [22]
[23]
[24]
[25]
[26]
Determination of the 235U content in uranium oxide particles by fission track analysis, Nucl. Instrum. Meth. Phys. Res. A 525 (2004) 582–592. C.G. Lee, K. Iguchi, F. Esaka, M. Magara, S. Sakurai, K. Watanabe, S. Usuda, Influence of uranium enrichment on the etching rate of polycarbonate fission track detector containing uranium particles, Nucl. Instrum. Methods Phys. Res. B 245 (2006) 440–444. M. Kraiem, S. Richter, H. Kühn, E.A. Stefaniak, G. Kerckhove, J. Truyens, Y. Aregbe, Investigation of uranium isotopic signatures in real-life particles from a nuclear facility by thermal ionization mass spectrometry, Anal. Chem. 83 (2011) 3011–3016. J.-H. Park, S. Park, K. Song, Isotopic analysis of NUSIMEP-6 uranium particles using SEM-TIMS, Mass Spectrom. Lett. 4 (2013) 51–54. F. Esaka, K. Watanabe, H. Fukuyama, T. Onodera, K.T. Esaka, M. Magara, S. Sakurai, S. Usuda, Efficient isotope ratio analysis of uranium particles in swipe samples by total-reflection X-ray fluorescence spectrometry and secondary ion mass spectrometry, J. Nucl. Sci. Technol. 41 (2004) 1027–1032. G. Tamborini, SIMS analysis of uranium and actinides in microparticles of different origin, Microchim. Acta 145 (2004) 237–242. F. Esaka, M. Magara, C.G. Lee, S. Sakurai, S. Usuda, N. Shinohara, Comparision of ICP-MS and SIMS techniques for determining uranium isotope ratios in individual particles, Talanta 78 (2009) 290–294. F. Esaka, M. Magara, D. Suzuki, Y. Miyamoto, C.G. Lee, T. Kimura, Isotope ratio analysis of individual sub-micrometer plutonium particles with inductively coupled plasma mass spectrometry, Talanta 83 (2010) 569–573. J. Ignacio, G. Alonso, Determination of fission products and actinides by inductively coupled plasma-mass spectrometry using isotope dilution analysis: a study of random and systematic errors, Anal. Chim. Acta 312 (1995) 57–78. M. Kraiem, S. Richter, N. Erdmann, H. Kühn, M. Hedberg, Y. Aregbe, Characterizing uranium oxide reference particles for isotopic abundances and uranium mass by single particle isotope dilution mass spectrometry, Anal. Chem. Acta 748 (2012) 37–44. J.-H. Park, S. Park, K. Song, Determination of the concentration and isotope ratio of uranium in soil and water by thermal ionization mass spectrometry, Mass. Spectom. Lett. 5 (2014) 12–15. D. Suzuki, Y. Saito-Kokubu, S. Kakurai, C.-G. Lee, M. Magara, K. Iguchi, T. Kimura, A new method for isotope ratio measurement of uranium in trace amount by thermal ionization mass spectrometry: the continuous heating method, Int. J. Mass Spectrom. 294 (2010) 23–27. J.-H. Park, I. Choi, S. Park, M. Lee, K. Song, A correction method for the peak tailing backgrounds for accurate isotope ratio measurements of uranium in ultra trace levels using thermal ionization mass spectrometry, Bull. Korean Chem. Soc. 32 (2011) 4327–4330. International Organization for Standardization (ISO), Guide to the Expression of Uncertainty in Measurement, ISBN 92-6r-r10188-9, 1995. S. Bürger, R.M. Essex, K.J. Mathew, S. Richter, R.B. Thomas, Implementation of guide to the expression of uncertainty in measurement (GUM) to multi-collector TIMS uranium isotope ratio metrology, Int. J. Mass Spectrom. 294 (2010) 65–76. International Atomic Energy Agency (IAEA), Qualification Procedure for the Network of Analytical laboratories for environmental sampling, SG-SGAS9006, 2011. V.A. Pavlenko, N.S. Pliss, B.N. Sokolov, A.P. Shcherbakov, Ion scattering in mass spectrometry. Theory and its application for abundance sensitivity estimation, Int. J. Mass Spectrom. Ion-. Phys. 46 (1983) 55–58. P. Calsteren, J.B. Schwieters, Performace of a thermal ionization mass spectrometer with a deceleration lens system and post-deceleration detector selection, Int. J. Mass Spectrom. Ion-. Process. 146–148 (1995) 119–129. S. Richter, S.A. Goldberg, Improved techniques for high accuracy isotope ratio measurements of nuclear materials using thermal ionization mass spectrometry, Int. J. Mass Spectrom. 229 (2003) 181–197.
Contents lists available at ScienceDirect
Talanta journal homepage: www.elsevier.com/locate/talanta
Simultaneous determination of the quantity and isotopic ratios of uranium in individual micro-particles by isotope dilution thermal ionization mass spectrometry (ID-TIMS) Jong-Ho Park n, Eun-Ju Choi Nuclear Chemistry Research Division, Korea Atomic Energy Research Institute, 111 Daedeok-daero-989, Yuseong-gu, Daejeon, 34057 Republic of Korea
art ic l e i nf o
a b s t r a c t
Article history: Received 14 July 2016 Received in revised form 27 July 2016 Accepted 1 August 2016 Available online 2 August 2016
A method to determine the quantity and isotopic ratios of uranium in individual micro-particles simultaneously by isotope dilution thermal ionization mass spectrometry (ID-TIMS) has been developed. This method consists of sequential sample and spike loading, ID-TIMS for isotopic measurement, and application of a series of mathematical procedures to remove the contribution of uranium in the spike. The homogeneity of evaporation and ionization of uranium content was confirmed by the consistent ratio of n(233U)/n(238U) determined by TIMS measurements. Verification of the method was performed using U030 solution droplets and U030 particles. Good agreements of resulting uranium quantity, n (235U)/n(238U), and n(236U)/n(238U) with the estimated or certified values showed the validity of this newly developed method for particle analysis when simultaneous determination of the quantity and isotopic ratios of uranium is required. & 2016 Elsevier B.V. All rights reserved.
Keywords: Uranium Thermal Ionization Mass Spectrometry Isotopic Ratio IDMS Particle Analysis
1. Introduction Reliable determination of the quantity and isotopic ratios of nuclear materials, such as uranium (U) and plutonium (Pu), in environmental samples plays an important role in monitoring undeclared nuclear activities for nuclear safeguards and in tracking the origin of samples for nuclear forensics [1–3]. Information on the amounts of U and/or Pu in a sample can be suggestive of deviation from declared activities, while the nuclear history of samples and the corresponding facilities can be tracked using the isotopic ratios. Therefore, the acquisition of both types of information is important for making accurate decisions relating nuclear safeguards. Nuclear material analysis can be classified into two main categories: bulk analysis and particle analysis. The objective of the former is to determine both the amounts and the isotopic ratios of nuclear materials in samples [4–6]. Because all of the nuclear particles in a sample are chemically treated and dissolved in a solution prior to isotopic measurement by inductively coupled plasma mass spectrometry (ICP-MS) or thermal ionization mass spectrometry (TIMS), the isotopic ratios acquired by bulk analysis are just average values. Therefore, the information obtained by bulk analysis provides only a rough clue to the occurrence of n
Corresponding author. E-mail address: [email protected] (J.-H. Park).
http://dx.doi.org/10.1016/j.talanta.2016.08.006 0039-9140/& 2016 Elsevier B.V. All rights reserved.
undeclared nuclear activities. The isotopic ratios of U and/or Pu in individual micro-particles contained in a sample can be determined by TIMS, which provides essential information with which to track the nuclear history of a facility. TIMS combined with the fission track (FT) technique or energy-dispersive spectroscopy in a scanning electron microscope (SEM-EDS) for particle identification is one of the best options to measure the isotopic ratios with high accuracy, precision, and sensitivity [7–11]. Secondary ion mass spectrometry (SIMS) is another method for particle analysis with a fast particle search process and reliable isotopic analysis [12–14]. However, the amount of nuclear materials is simply estimated by measuring the particle size, with the assumption that the micro-particles consist purely of U or Pu and are spherical in shape. When particles deviate from these assumptions, there is a similar deviation of the determined quantity from the true value. Recently, ICP-MS has been utilized for isotopic ratio analysis of individual sub-micrometer plutonium particles by a combination of chemical dissolution and separation processes [15]. Quantitative analysis of Pu is possible by measuring the signal intensity and by utilizing the standard curve. This method is useful when sufficient instrumental sensitivity is ensured and a negligible background from the chemical process is guaranteed. However, quantitative and isotopic ratio analyses by this method are hardly available for individual micro-particles of uranium, because a considerable amount of U background is expected during the chemical process. Isotope dilution mass spectrometry (IDMS) is the most accurate
J.-H. Park, E.-J. Choi / Talanta 160 (2016) 600–606
technique for quantitative analysis [16–18]. In this approach, a known amount of spike reference solution is mixed with a sample solution followed by isotopic analysis using ICP-MS or TIMS. This technique is limited, however, in cases when a sample can hardly be dissolved [17] due to inhomogeneous mixing with the spike. Determination of the quantity of nuclear micro-particles is included in such cases. Furthermore, mixing a sample with the spike results in distortion of the isotopic ratios of the target element. The sample solution is normally divided into two portions: one for mixing with the spike reference solution and the other for isotopic ratio determination. However, if a sample consists of micro-particles, this process is not possible, which means that an additional technique to remove the distortion of the isotopic ratio is required. Kraiem et al. used isotope dilution thermal ionization mass spectrometry (ID-TIMS) to characterize uranium particles, which were monodisperse uranium oxide particles of a known size (ca. 1 mm), from a certified uranium standard solution [17]. The applicability of the aforementioned technique for quantitative analysis of uranium particles was successfully verified, and the determination of uranium isotopic ratios was also performed. However, the quantitative and isotopic analyses were carried out using monodisperse duplicated particles. If the isotopic ratios of U and/ or Pu in individual particles are expected not to be identical, as in the case of typical environmental samples, only one of quantitative and isotopic analyses can be performed, which means that a new method to perform both is required. This paper presents a method to determine both quantity and isotopic ratios of uranium contained in individual micro-particles by ID-TIMS. Homogeneous evaporation and ionization of the uranium from a sample and a spike are described for verification of the application of IDMS to the quantification of uranium in microparticles. A mathematical model and a reasonable assumption were applied to remove the contribution of the spike to the isotopic ratios of uranium in individual micro-particles. The quantitative and isotopic analytical results are presented to validate the newly developed method for particle analysis.
2. Experimental 2.1. Spike and sample materials IRMM-040a (EU-JRC-IRMM), as the certified spike isotopic reference material, was diluted with nitric acid (2 M) to make a spike solution. The amount of 233U in the spike solution was 2.065 2 [ 70.003 6] 10 10 mol g 1. (The numbers in parentheses in this paper indicate expanded uncertainty, Ue ¼ k uc, where k and uc are the coverage factor and the combined standard uncertainty, respectively.). U030 (National Bureau of Standards, USA), a certified reference material, was used for the test samples to validate the test method. The test samples were divided into two main types: nitric acid (2 M) solution samples of U030 (called ‘U030 solution’) and microparticle samples of U030 powder (highly purified oxide, U3O8). The amount of 238U in the U030 solution was determined preliminarily to be 2.261 [7 0.014] 10 10 mol g 1 by IDMS. 2.2. Sample-spike mixture preparation Sample loading and spike addition were sequentially performed on a zone-refined rhenium filament (Thermo Fisher Scientific, Bremen, Germany). Firstly, 1 μL of U030 solution was transferred onto a filament using a micro-pipette; the relative uncertainty of this volume was 1.5%. The calculated amount of 238U in the transferred U030 solution was 57.2 [ 71.8] pg. Once the transferred U030 solution was completely dried by heat created by
601
a current of 0.6 A, 1 μL of the spike solution was transferred onto the same spot where the dried U030 solution was located, and was then dried. Our preliminary experiment showed that mixing by the aforementioned process resulted in inhomogeneous evaporation and ionization of 233U and 238U, and the signal intensity ratio of n(233U)/n(238U) was not constant throughout a measurement. To resolve this problem, re-dissolution of the mixture was performed by adding 1 M HNO3 on the spot where the mixture was located, followed by drying. Two volumes (i.e., 2 μL) of nitric acid were added to cover the spot sufficiently for complete re-dissolution of the mixture. Re-dissolution was repeated five times, and then the mixture was fixed on the filament by applying a current of 1.8 A for 30 s; five samples were prepared (S1–S5). In the case of micro-particle samples, a micro-manipulator system equipped in a scanning electron microscope (SEM) was utilized to transfer a single particle (ca. 2 mm) to a filament (Fig. 1). The rest of the procedure was the same as for the case of the U030 solution; five samples were prepared (P1–P5). 2.3. Isotopic measurement and data analysis The isotopic measurement using TIMS (Triton Plus; Thermo Fisher Scientific) adopting the continuous heating method was performed as described elsewhere [6,11,18–20]. A multi-ion counter system, consisting of three secondary electron multipliers (SEMs) and two compact discrete dynode detectors (CDDs), was used for simultaneous measurement. The detector configuration used in this study is described in Table 1. Detection efficiencies for the five ion counters were adjusted using the ion signals of 187Re (ca. 300,000 cps). The mass bias for the isotopes, with the exception of 233U, was corrected by isotopic measurement of a reference material (U200; National Bureau of Standards, USA). The mass bias for 233U was not corrected due to the lack of availability of an appropriate reference material. Only the data sets, whose 238U þ intensity fell in the range of 5– 100% with respect to the maximum 238U þ intensity, were considered as the valid data sets to avoid distortion of the resulting isotopic ratios and unnecessarily large uncertainty arising from relatively small signal intensity. The weights for the individual data sets were considered to calculate the weighted means and the weighted standard errors of the isotopic measurements. The uncertainty was estimated according to GUM (Guide to the Expression of Uncertainty in Measurement), in compliance with ISO/IEC Guide 98-3 [21,22]. All of the analytical processes were performed in a clean facility, which was controlled to ISO 5 (class 100) and ISO 6 (class 1000) levels to avoid sample contamination.
3. Results and discussion 3.1. Homogeneous evaporation and ionization of uranium The mass spectrometric behavior, as well as the chemical behavior of 233U from the spike and 238U from a sample, must be identical for valid application of IDMS. Figs. 2(a) and 3(b) are the examples of 233U þ and 238U þ intensity profiles during TIMS measurements for a mixture of a U030 solution droplet and the spike, and a mixture of a U030 particle and the spike, respectively. The dash-dotted lines indicate the 5% levels of the maximum intensities of 238U þ in each profiles, and the valid data sets are above them. The profiles shows that the evaporation and ionization of the two isotopes are identical during a measurement. The change in the isotopic ratios of n(233U)/n(238U) in the range of valid data sets, as shown in Figs. 2(b) and 3(b), were approximately 5–8%. This amount of change is typical when the mass fractionation of
602
J.-H. Park, E.-J. Choi / Talanta 160 (2016) 600–606
Fig. 1. Micro-particle transfer to a filament using a micromanipulation system.
TIMS is considered, and is not responsible for the different mass spectrometric behavior of 233U and 238U. Therefore, although homogeneous mixing of the spike and a sample was not guaranteed, no inhomogeneity in evaporation and ionization of the two isotopes was observed.
Table1 Detector configuration of TIMS in this study. Isotope
234
IC4 CDD
IC3 SEM
U
Name Type
Ion Signal Intensity (cps)
Detector
233
U
235
236
238
IC2 SEM
IC1 SEM
IC5 CDD
U
U
U
800.0k
(a) 600.0k U U
400.0k 200.0k
5%
0.0 0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
1.00
n(
233
U)/n(
238
U)
0.95
(b)
0.90 0.85 0.80 0.75 0.70 0.65 2600
2800
3000
3200
3400
3600
3800
4000
Current Applied to Evaporational Filament (mA) Fig. 2. The profiles of 233U þ and 238U þ intensities during TIMS measurement for a mixture of a U030 solution droplet and the spike (a) and the change in n(233U)/n(238U) in the range of valid data sets (b). The dash-dotted line in (a) indicates the 5% level of the maximum intensity of 238U þ to determine the range of valid data sets.
J.-H. Park, E.-J. Choi / Talanta 160 (2016) 600–606
603
Fig. 3. The profiles of 233U þ and 238U þ intensities during TIMS measurement for a mixture of a U030 particle and the spike (a) and the change in n(233U)/n(238U) in the range of valid data sets (b). The dash-dotted line in (a) indicates the 5% level of the maximum intensity of 238U þ to determine the range of valid data sets.
3.2. Calculation of quantity and isotopic ratios The amount of 238U in a sample is determined by the following equations [15–17]:
(
csample
238
)
U =
rspike − rmixture 1 + rsample Vspike⋅dspike ⋅ ⋅ rmixture − rsample 1 + rspike Vsample⋅dsample
(
⋅cspike
233
U
)
(
238
(
238
csample
(1)
or
asample
can be completely ruled out, for example, by periodic monitoring of the spike solution in compliance with a quality assurance and quality control program. Under this assumption, Eqs. (1) and (2) are reduced as follows, respectively:
(
)
(
233
)
U ,
(2)
where csample(238U) and cspike(233U) are the amounts content (mol g 1) of 238U in a sample and 233U in the spike solution; asample(238U) and aspike(233U) are the amounts (mol) of 238U in a sample and 233U in the spike solution added to a sample, respectively; rsample, rspike, and rmixture are the amount ratio n(233U)/n (238U) in a sample, the spike, and a mixture, respectively; Vsample and Vspike are the volumes of a sample and the spike, respectively; and dsample and dspike are the densities of a sample and the spike, respectively. The amounts of 238U and 233U [i.e., asample(238U) and aspike(233U), respectively] are the products of the corresponding volume (Vsample and Vspike), density (dsample and dspike), and the amount contents [csample(238U) and cspike(233U)], respectively, and contain their propagated uncertainties. To acquire the three isotopic ratios (rsample, rspike, and rmixture), the uranium isotopic measurement by TIMS must be performed for a sample, the spike, and a mixture in normal cases. However, the measurement for a sample to obtain rsample is not possible in this case as no replica is available. Therefore, the following assumption is necessary for application of ID-TIMS to the proposed method: no 233U nuclide is contained in a sample, which enables us to set rsample to zero. Because 233U is a fissile isotope of uranium produced by the neutron irradiation of 232Th (thorium-232), its natural abundance is negligible. Therefore, this assumption is reasonable. In addition, for practical purposes, the certified value of the spike material can be used for the amount ratio n(233U)/n (238U) in the spike (i.e., rspike) if the possibility of contamination
rspike − rmixture rmixture
⋅
Vspike⋅dspike 1 ⋅ ⋅cspike 1 + rspike Vsample⋅dsample
(
233
U
)
(3)
or
asample rspike − rmixture 1 + rsample 238 U = ⋅ ⋅aspike rmixture − rsample 1 + rspike
)
U =
)
U =
rspike − rmixture rmixture
⋅
1 ⋅aspike 1 + rspike
(
233
)
U .
(4)
The isotopic ratios of n(234U)/n(238U), n(235U)/n(238U), and n (236U)/n(238U) in a mixture (Rmixture) are the combination of the corresponding ratios of uranium in a sample (Rsample) and the spike (Rspike) with a contribution factor (s). Therefore, Rsample was determined as follows:
Rsample =
R mixture − (1 − s )⋅Rspike (5)
s
where
s=
asample(238U ) a mixture(
238
U)
aspike(238U )
=1−
a mixture(238U )
The total amount of culated using Eq. (7).
.
238
(6) 238
U in a mixture [amixture(
U)] was cal-
a mixture(238U ) = asample(238U ) + aspike(238U ) = asample(238U ) + aspike(233U )⋅
1 rspike
(7)
The amount and the amount content of total uranium in a sample [asample(U) and csample(U), respectively] are determined based on the amount and the amount content of 238U, and the isotope ratios calculated above as follows:
asample(U ) = asample
(
238
)(
U ⋅ 1+
(
csample(U ) = csample(238U )⋅ 1 +
∑ Rsample),
∑ Rsample).
(8) (9)
604
J.-H. Park, E.-J. Choi / Talanta 160 (2016) 600–606
Table 2 The amount of total U and the isotopic ratios of U in the U030 solution droplets.
Estimated Value Certified Value S1 S2 S3 S4 S5 Mean value Relative Unc.a (%) Accuracyb (%) RSDc (%)
Total U Isotopic Ratio amount (pg) n (234U)/n (238U) ( 10 4)
n (235U)/n (238U) ( 10 2)
n (236U)/n (238U) ( 10 4)
57.2 [ 7 1.8]
–
–
–
–
1.961 [7 0.010] 1.84 [ 7 0.57] 1.55 [ 7 0.91] 1.77 [ 70.42] 1.86 [ 7 0.45] 2.12 [ 7 0.45] 1.83 31
3.143 0 [7 0.003 0] 3.16 [ 70.21] 3.13 [ 70.21] 3.13 [ 70.21] 3.13 [ 70.21] 3.12 [ 7 0.21] 3.13 6.7
2.105 [ 70.010] 2.21 [ 7 0.18] 2.20 [ 7 0.16] 2.21 [ 7 0.15] 2.21 [ 7 0.15] 2.20 [ 7 0.15] 2.21 7.2
6.8 11
0.34 0.50
9.4 0.33
53.5 54.4 60.6 56.8 56.5 56.4 3.0
[7 1.7] [7 1.6] [7 1.9] [7 1.7] [7 1.7]
1.5 4.9
The numbers in parentheses indicate expanded uncertainty, Ue ¼ k uc , where uc is the combined uncertainty and k ¼ 2 with 95% confidential level. a Relative Unc. is the relative uncertainty defined as (Uncertainty)/(Mean Value)*100. b Accuracy is defined as (Measure value – Certified value)/(Certified value) *100. c RSD is the relative standard deviation.
3.3. Determination of quantity and isotopic ratios of U in U030 solution samples Table 2 and Fig. 4 show the amount of total U and the isotopic ratios of U contained in the individual U030 solution droplets (ca. 64 62
1 μL) determined by the proposed method. The estimated amount of transferred uranium on a filament (estimated value) and the certified isotopic ratios (certified value) are also shown. The uncertainty ranges of the U quantity determined by IDTIMS overlapped with the estimated value. The accuracy ranged from 6.5% to 6.0%, the average of which was 1.5%. The relative standard deviation (RSD) of the resulting amounts was 4.9% and only one measurement (S1) deviated slightly from the uncertainty range of the estimated value. Although the performance of the quantification by the proposed method was not as accurate and precise as that by normal IDMS, it was still acceptable for determination of the amount of U, considering that the amount is at an ultra-trace level (ca. 50 pg). For example, the measurement goal of the relative expanded uncertainty of U total mass for the analysis of bulk environmental samples needs to be less than 40% [23]. The isotopic ratios of n(235U)/n(238U) and n(236U)/n(238U) were fairly consistent with the corresponding certified values within the uncertainty ranges. The deviation of n(236U)/n(238U) was considered to be mostly due to the peak tailing contribution of 238U to 236 U, because positive bias was consistently observed [24–26]. The RSDs of n(235U)/n(238U) and n(236U)/n(238U) were 0.50% and 0.30%, respectively, indicating high repeatability. The isotopic ratio of n (234U)/n(238U) were also consistent with the certified value. However, the determination was not highly reliable since the relative uncertainty and RSD were somewhat high (31% and 11%, respectively). This is because the spike material used in this study contained a considerable amount of 234U (ca. 0.93% of 233U), which blocks the contribution of 234U from a sample when mixed for IDTIMS. As shown in Eq. (5), Rmixture is determined mainly by Rspike and the contribution of Rsample becomes minor when Rspike is considerably greater than Rsample. Determination of n(234U)/n(238U) is included in such case because Rspike is 1.137 287 0.000 24 and -4
2.8x10
(a)
-4
2.6x10
(b)
-4
60
2.4x10 Upper Range
-4
2.2x10
58
-4
2.0x10
-4
1.8x10
56 Lower Range
-4
1.6x10 54
-4
1.4x10
-4
52
U Qauntity (pg)
-2
-4
3.0x10
(c)
-4
234
238
236
238
n( U)/n( U)
-4
1.0x10
50 4.5x10
1.2x10
(d)
2.8x10
-2
4.0x10
-4
2.6x10
-4
2.4x10
-2
3.5x10
-4
2.2x10 -2
3.0x10
-4
2.0x10
-4
1.8x10
-2
2.5x10
-4
1.6x10 -2
235
2.0x10
238
n( U)/n( U) S1
S2
S3
S4
S5
-4
1.4x10
n( U)/n( U)
-4
1.2x10
S1
S2
S3
S4
S5
Fig. 4. The amount of total U (a) and the isotopic ratios of n(234U)/n(238U) (b), n(235U)/n(238U) (c), n(236U)/n(238U) (d) contained in the U030 solution samples. The dashed line and the dash-dotted lines in (a) are the estimated value of U quantity and the upper/lower ranges of uncertainty of the estimated value, respectively. The dashed lines in (b), (c), and (d) are the certified values.
J.-H. Park, E.-J. Choi / Talanta 160 (2016) 600–606
Table 3 The amount of total U and the isotopic ratios of U in the U030 particles. Total U amount (pg)
P1 P2 P3 P4 P5 Mean value Relative Unc. (%) Accuracy (%) RSD (%)
Isotopic Ratio n (234U)/n (238U) ( 10 4)
n (235U)/n (238U) ( 10 2)
36.5 [7 1.1] 21.0 [ 7 0.9] 25.5 [7 1.1] 21.4 [ 7 0.8] 22.0 [7 1.1] – –
7.1 [ 7 1.2] 5.8 [ 7 7.9] 4.3 [ 7 6.1] 7.9 [7 5.0] 2.5 [ 7 8.9] 5.5 110
3.18 3.14 3.13 3.11 3.15 3.14 8.3
– –
181.4 39.5
0.08 0.81
n (236U)/n (238U) ( 10 4)
[ 7 0.21] [ 7 0.28] [ 7 0.27] [7 0.23] [ 7 0.31]
2.30 2.25 2.35 2.24 2.36 2.30 13
[ 70.18] [ 70.34] [7 0.28] [7 0.27] [ 70.37]
9.4 2.43
The numbers in parentheses indicate expanded uncertainty, Ue ¼ k uc , where uc is the combined uncertainty and k ¼ 2 with 95% confidential level.
Rsample is in the order of 10 4. The accurate value of n(235U)/n(238U) reconfirmed that the U quantity values determined by this method were acceptable because the isotopic ratios are closely related to the quantity information [see Eqs. (5)–(9)]. 3.4. Determination of quantity and isotopic ratios of U in U030 particles The method for the simultaneous determination of the quantity and isotopic ratios of uranium was applied to the particle samples (P1 to P5), the results of which are summarized in Table 3. Because the U030 particles in this study were not well shaped, the accurate estimation of the uranium amount by size measurement was not possible. Instead, roughly 30 pg of uranium (with a high level of uncertainty) was estimated to be contained in individual particles. The amounts of uranium in the U030 particles measured by the proposed method fell into a range close to the rough estimate. There is no direct way to verify whether the measurement results were consistent with real values, as in the case of U030 solution samples. However, indirect verification is possible by checking the isotopic ratios because the quantity information is related to these ratios. Agreement of the isotopic ratios of n(235U)/n(238U) and n(236U)/ n(238U) with the certified values within the uncertainty ranges was observed, as more clearly shown in Fig. 5. The accuracies of individual measurements for n(235U)/n(238U) and n(236U)/n(238U) ranged from 0.5% to 1.1% and from 6.4% to 12.0%, respectively. Although consistent positive bias was observed for n(236U)/n 4.5x10
-2
4.2x10
-2
3.9x10
-2
3.6x10
-2
3.3x10
-2
3.0x10
-2
2.7x10
-2
2.4x10
-2
2.1x10
-2
1.8x10
-2
235
(a)
238
n( U)/n( U)
P1
P2 235
Fig. 5. The isotopic ratios of n(
U)/n(
P3
238
P4 236
U) (a) and n(
U)/n(
238
P5
605
(238U), this was mostly due to the peak tailing effect as in the case of U030 solution samples, and is typically observed in TIMS measurements unless a correction method is applied or a retarding potential quadrupole (RPQ) system is utilized [25]. The relative uncertainty and RSD were 8.3% and 0.81%, respectively, for n (235U)/n(238U), and 13% and 2.43% for n(236U)/n(238U). On the other hand, the isotopic ratio of n(234U)/n(238U) deviated markedly from the certified values. The average accuracy, relative uncertainty, and RSD were 181.4%, 110%, and 39.5%, respectively. As explained above, the large deviation and low precision are caused by the high content of 234U in the spike, which blocked the 234U in the samples. Furthermore, this effect was relatively enhanced for the particle analysis compared with that for the U030 solution droplet analysis, because the amounts of 234U in individual particles were approximately half those in U030 solution droplet samples, while those in the spike were the same. However, selecting a more appropriate spike material, which contains an extremely small amount of 234U, would resolve this problem. Despite the inaccurate results for n(234U)/n(238U) due to the adoption of an inappropriate spike material, the determination of n(235U)/n(238U) and n(236U)/n(238U) using the proposed method is considered to be highly reliable. The analytical result of n(235U)/n (238U) satisfied the accuracy requirements for analysis of particle environmental samples relating to International Atomic Energy Agency (IAEA) safeguards (i.e., less than 1%) [23]. In addition, the accuracy of the isotopic ratio determination is strong evidence that the determination of U quantity was also reliable, as mentioned above. Based on the verification of the analytical performance, the newly developed method can be used for the analysis of uranium in individual particles with an acceptable level of measurement performance. This is especially useful when the quantity and isotopic ratios of uranium must be determined simultaneously and replicated particles are not available.
4. Conclusion A newly developed method to determine the quantity and isotopic ratios of uranium contained in individual micro-particles simultaneously by ID-TIMS is described here. While the determinations of the quantity and isotopic ratios are performed separately by IDMS and TIMS (or ICP-MS), respectively, in normal cases, particle analysis allows only one measurement when replicated particles are not available. Two types of sample (U030 solution droplets and particles) were used for verification of the method. A mixture for ID-TIMS was prepared by sequential loading of a sample and the spike on a TIMS filament, followed by repeated 3.0x10
-4
2.8x10
-4
2.6x10
-4
2.4x10
-4
2.2x10
-4
2.0x10
-4
1.8x10
-4
1.6x10
-4
1.4x10
-4
1.2x10
-4
(b)
P1
236
238
n( U)/n( U)
P2
P3
P4
P5
U) (b) contained in the U030 particles. The dashed lines in (b), (c), and (d) are the certified values.
606
J.-H. Park, E.-J. Choi / Talanta 160 (2016) 600–606
dissolution with nitric acid. Homogeneous evaporation and ionization of 233U and 238U contents were confirmed for validation of the application of ID-TIMS for this purpose. An additional series of mathematical processes was applied to remove the contribution of the spike to the isotopic ratios of uranium in a sample. Fair agreement with the estimated or certified values was observed for the quantity and the isotopic ratios of n(235U)/n(238U) and n(236U)/ n(238U) for the U030 solution droplet sample measurement, the accuracies of which were 1.5%, 0.34%, and 9.4%, respectively, while the reliability of n(234U)/n(238U) analysis was not ensured due to the large amount of 234U in the spike material. The application of this method to particle analysis was verified by the analytical results of the isotopic ratios of n(235U)/n(238U) and n (236U)/n(238U). The deviation and the large RSD in the n(234U)/n (238U) analysis due to the factor mentioned above can be resolved by selecting more appropriate spike materials that contain a lower level of 234U. The analytical method proposed here should be useful for micro-particle analysis when both the quantity and the isotopic ratio of uranium are required, but when no replicated particles are available.
[9]
[10]
[11] [12]
[13] [14]
[15]
[16]
[17]
Acknowledgments This work was supported by the Ministry of Science, ICT, and Future Planning (MSIP) and the Nuclear Safety and Security Commission (NSSC) of South Korea (No. 1405020).
[18]
[19]
References [20] [1] D.L. Donohue, Strengthening IAEA safeguards through environmental sampling and analysis, J. Alloy Compd. 271–273 (1998) 11–18. [2] D.L. Donohue, Strengthended nuclear safeguards, Anal. Chem. 74 (2002) 28A–35A. [3] K. Mayer, M. Wallenius, Z. Varga, Nuclear forensic science: correlating measurable material parameters to the history of nuclear material, Chem. Rev. 113 (2013) 884–900. [4] M. Magara, Y. Hanzawa, F. Esaka, Y. Miyamoto, K. Yasuda, K. Watanabe, S. Usuda, H. Nishimura, T. Adachi, Development of analytical technique for ultra trace amounts of nuclear materials in environmental samples using ICPMS for safeguards, Appl. Radiat. Isot. 53 (2000) 87–90. [5] O.T. Farmer III, K.B. Olsen, M.L. Thomas, S.J. Garofoli, Analysis of IAEA environmental samples for plutonium and uranium by ICP/MS in support of international safeguards, J. Radioanal. Nucl. Chem. 276 (2008) 489–492. [6] J.-H. Park, S. Lee, Y.-G. Ha, S.A. Lee, K. Jeong, M. Lee, K. Song, Bulk analysis of a simulated environmental sample in natural abundance performed in KAERI for nuclear safeguards, J. Radioanal. Nucl. Chem. 303 (2015) 1297–1300. [7] K.G. Heumann, S. Eisenhut, S. Gallus, E.H. Hebeda, R. Nusko, A. Vengosh, T. Walczyk, Recent developments in thermal ionization mass spectrometric techniques for isotope analysis, Analyst 120 (1995) 1291–1299. [8] O. Stetzer, M. Betti, J. Geel, N. Erdmann, J.-V. Kratz, R. Schenkel, N. Trautmann,
[21] [22]
[23]
[24]
[25]
[26]
Determination of the 235U content in uranium oxide particles by fission track analysis, Nucl. Instrum. Meth. Phys. Res. A 525 (2004) 582–592. C.G. Lee, K. Iguchi, F. Esaka, M. Magara, S. Sakurai, K. Watanabe, S. Usuda, Influence of uranium enrichment on the etching rate of polycarbonate fission track detector containing uranium particles, Nucl. Instrum. Methods Phys. Res. B 245 (2006) 440–444. M. Kraiem, S. Richter, H. Kühn, E.A. Stefaniak, G. Kerckhove, J. Truyens, Y. Aregbe, Investigation of uranium isotopic signatures in real-life particles from a nuclear facility by thermal ionization mass spectrometry, Anal. Chem. 83 (2011) 3011–3016. J.-H. Park, S. Park, K. Song, Isotopic analysis of NUSIMEP-6 uranium particles using SEM-TIMS, Mass Spectrom. Lett. 4 (2013) 51–54. F. Esaka, K. Watanabe, H. Fukuyama, T. Onodera, K.T. Esaka, M. Magara, S. Sakurai, S. Usuda, Efficient isotope ratio analysis of uranium particles in swipe samples by total-reflection X-ray fluorescence spectrometry and secondary ion mass spectrometry, J. Nucl. Sci. Technol. 41 (2004) 1027–1032. G. Tamborini, SIMS analysis of uranium and actinides in microparticles of different origin, Microchim. Acta 145 (2004) 237–242. F. Esaka, M. Magara, C.G. Lee, S. Sakurai, S. Usuda, N. Shinohara, Comparision of ICP-MS and SIMS techniques for determining uranium isotope ratios in individual particles, Talanta 78 (2009) 290–294. F. Esaka, M. Magara, D. Suzuki, Y. Miyamoto, C.G. Lee, T. Kimura, Isotope ratio analysis of individual sub-micrometer plutonium particles with inductively coupled plasma mass spectrometry, Talanta 83 (2010) 569–573. J. Ignacio, G. Alonso, Determination of fission products and actinides by inductively coupled plasma-mass spectrometry using isotope dilution analysis: a study of random and systematic errors, Anal. Chim. Acta 312 (1995) 57–78. M. Kraiem, S. Richter, N. Erdmann, H. Kühn, M. Hedberg, Y. Aregbe, Characterizing uranium oxide reference particles for isotopic abundances and uranium mass by single particle isotope dilution mass spectrometry, Anal. Chem. Acta 748 (2012) 37–44. J.-H. Park, S. Park, K. Song, Determination of the concentration and isotope ratio of uranium in soil and water by thermal ionization mass spectrometry, Mass. Spectom. Lett. 5 (2014) 12–15. D. Suzuki, Y. Saito-Kokubu, S. Kakurai, C.-G. Lee, M. Magara, K. Iguchi, T. Kimura, A new method for isotope ratio measurement of uranium in trace amount by thermal ionization mass spectrometry: the continuous heating method, Int. J. Mass Spectrom. 294 (2010) 23–27. J.-H. Park, I. Choi, S. Park, M. Lee, K. Song, A correction method for the peak tailing backgrounds for accurate isotope ratio measurements of uranium in ultra trace levels using thermal ionization mass spectrometry, Bull. Korean Chem. Soc. 32 (2011) 4327–4330. International Organization for Standardization (ISO), Guide to the Expression of Uncertainty in Measurement, ISBN 92-6r-r10188-9, 1995. S. Bürger, R.M. Essex, K.J. Mathew, S. Richter, R.B. Thomas, Implementation of guide to the expression of uncertainty in measurement (GUM) to multi-collector TIMS uranium isotope ratio metrology, Int. J. Mass Spectrom. 294 (2010) 65–76. International Atomic Energy Agency (IAEA), Qualification Procedure for the Network of Analytical laboratories for environmental sampling, SG-SGAS9006, 2011. V.A. Pavlenko, N.S. Pliss, B.N. Sokolov, A.P. Shcherbakov, Ion scattering in mass spectrometry. Theory and its application for abundance sensitivity estimation, Int. J. Mass Spectrom. Ion-. Phys. 46 (1983) 55–58. P. Calsteren, J.B. Schwieters, Performace of a thermal ionization mass spectrometer with a deceleration lens system and post-deceleration detector selection, Int. J. Mass Spectrom. Ion-. Process. 146–148 (1995) 119–129. S. Richter, S.A. Goldberg, Improved techniques for high accuracy isotope ratio measurements of nuclear materials using thermal ionization mass spectrometry, Int. J. Mass Spectrom. 229 (2003) 181–197.