Spatial distribution of uranium isotopes in solid

Microchemical Journal 140 (2018) 24–30

Contents lists available at ScienceDirect

Microchemical Journal journal homepage: www.elsevier.com/locate/microc

Spatial distribution of uranium isotopes in solid nuclear materials using laser ablation multi-collector ICP-MS Michael Krachler ⁎, Zsolt Varga, Adrian Nicholl, Maria Wallenius, Klaus Mayer European Commission - Joint Research Centre, Directorate for Nuclear Safety and Security, P.O. Box 2340, D-76125 Karlsruhe, Germany

a r t i c l e

i n f o

Article history: Received 19 March 2018 Received in revised form 29 March 2018 Accepted 29 March 2018 Available online 30 March 2018 Keywords: ICP-MS Laser ablation Uranium isotopes Enrichment Homogeneity Nuclear

a b s t r a c t Employing laser ablation multi-collector ICP-MS (LA-MC-ICP-MS), its potential for evaluating the homogeneity of solid uranium (U) bearing materials was investigated. To this end, the n(235U)/n(238U) ratio of two low-enriched U certified reference materials (CRMs), i.e. powdered standard reference material U-010 (~1 wt% 235U) and a UO2 pellet of CRM 125-A (~4 wt% 235U) was determined using line scan analysis. Four spots of 5 μm diameter each were ablated per second with the LA system moved at a speed of 20 μm s−1 for several minutes. Experimental and certified U isotope ratios matched perfectly, unequivocally validating the accuracy of the applied analytical methodology. In addition, the narrow and symmetric frequency distribution of the n(235U)/n(238U) ratio of both CRMs confirmed their homogeneity with respect to the U isotopic composition. Utilizing LA-MC-ICP-MS line scan analysis of two similar UO2 pellets (~1 wt% 235U) from the 5th Collaborative Materials Exercise (CMX-5) revealed diverse inhomogeneity regarding their n(235U)/n(238U) ratio. Although both UO2 pellets used for the CMX-5 exercise were prepared from identical source materials, their different production routes yielded largely contrasting frequency distributions of the n(235U)/n(238U) ratio. While 235U isotope abundance ranged from 0.75%–1.6% for the first pellet, this value fluctuated between 0.45% and 3.0% for the second pellet. This specific information sheds additional light on the production process of these materials and is helpful in nuclear forensics investigations when determining the origin of unknown nuclear material. The depleted nature (~0.4 wt% 235U) of two seized U metal samples was established through LA-MC-ICP-MS analysis of the n (235U)/n(238U) ratio. As black and yellow regions within both U bearing samples had identical U isotopic composition, the observed colour differences might arise from different degrees of U oxidation and are not associated with U isotopic inhomogeneity within the material. The significantly different 235U abundance of the two samples (0.3670 ± 0.0015 wt.% and 0.4146 ± 0.0013 wt.%), however, clearly indicates they were prepared from different source materials and/or from diverse amounts of them. © 2018 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction Nuclear forensic science, referred to as nuclear forensics, aims at identifying the origin of nuclear material found out of regulatory control [1]. More specifically, scientific analysis, focussing on nuclear or other radioactive material, or on other evidence that is contaminated with radioactive material, seeks for clues on its intended use and history [2]. Nuclear forensics additionally plays an important role in nuclear security regimes to prevent and respond to nuclear security events such as illicit trafficking of nuclear material. To this end, nuclear forensics utilizes material properties, so-called signatures that are inherent to the material and help revealing its processing history [1,2]. In this context, several measurable parameters are considered, such as isotope abundance ratios of lead, strontium, neodymium, sulphur, plutonium and uranium (U), the concentration of particular trace elements, as well as ⁎ Corresponding author. E-mail address: [email protected]. (M. Krachler).

daughter products of radioactive decay used for radiometric age dating [1–4]. In nuclear forensic investigations the determination of U isotopic signatures plays a key role. Although the minor U isotopes 234U and 236 U also carry meaningful information, the n(235U)/n(238U) ratio provides the level of U enrichment [1,2,5,6] which is, in many cases, the key parameter for decision if the possession of the material is consistent with legal requirements (e.g. natural U, low-enriched U or highly enriched U). Besides gamma spectrometry and thermal ionization mass spectrometry (TIMS), inductively coupled plasma-mass spectrometry (ICP-MS) is increasingly being applied for U isotopic analysis. While time-of-flight [7] and single collector [5,8–15] mass spectrometers have been employed repeatedly, ICP-MS instruments with multiple detectors steadily gain attraction for this purpose [6,15–20]. Also in nuclear forensics, laser ablation (LA) may serve as a powerful tool for sample introduction into the ICP-MS [16–20]. As only very small portions of the sample are required for LA-based analysis, the evidence, i.e. the solid sample, is available for further destructive analysis or can be archived

https://doi.org/10.1016/j.microc.2018.03.038 0026-265X/© 2018 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

M. Krachler et al. / Microchemical Journal 140 (2018) 24–30

for future analysis. Furthermore, the measurement can be completed rapidly (within a few hours), thus providing quick information to the investigating authority. Irrespective of the measurement technique employed for analysis, the correct interpretation of the experimental data and the final conclusions rely on high quality analytical results. The regular participation in inter-laboratory comparison programmes (ILCPs) is beneficial to check the performance of the enrolled laboratories for a specific measurement parameter and helps detecting weak points in the applied analytical process. Our laboratory has been taking part in various ILCP for N20 years, further demonstrating the accuracy of its analytical data and fitness for purpose of the analytical procedures. The Nuclear Forensics International Technical Working Group (ITWG) has repeatedly organized Collaborative Materials Exercises (CMXs) on sample materials such as plutonium oxide powder, low and highly enriched U as metal, powder or pellets [15,21,22]. These samples reflect the types of specimens that are frequently dealt with in a nuclear forensics laboratory. While CMXs are not intended to serve as proficiency tests for the participating laboratories, they were developed in order to provide a common problem set from which nuclear forensic techniques and methods can be compared and prioritized, establishing consensus decisions by the CMX participants. The fifth and most recent exercise of this series, CMX-5, involved two low-enriched uranium oxide (UO2) pellets [23]. The accurate determination of the U isotopic composition of each of these pellets was one of the requested parameters. As initial results needed to be reported within 24 h after receipt of the test samples, a first screening of the abundances of 234U, 235U, and 238U was performed on the entire solid samples using high resolution gamma spectrometry. These measurements provided a first helpful indication that the abundance of 235U of each of the two UO2 pellets is slightly enriched (~1 wt%). Two restrictions, however, limit the significance of this initial finding. First, the combined relative uncertainty of the initial results (obtained by gamma spectrometry), amounting to about 15%, was relatively high. Secondly, as the entire sample is measured, only the average U isotopic composition of the pellets was obtained. The gamma spectrometry results are only meaningful, if the investigated pellets are homogeneous with respect to their U isotopic composition. The main aim of the present study was the development and validation of an alternative “non-destructive” procedure for the accurate, precise and rapid determination of the U isotopic composition of U materials using LA-MC-ICP-MS. Results presented here highlight the significance of spatial resolution of such analysis that unmistakably reveal the (in-) homogeneity of the U isotopic composition of pressed U powder, UO2 pellets and fragments of U metal.

25

other samples within the LA chamber. The generated plume was transported from the LA cell to the MC-ICP-MS plasma using a laminar flow of Ar gas. Further details on the optimized instrumental settings as well as on the applied data acquisition parameters are reported elsewhere [20]. 2.2. Reagents and standards High-purity water (18.2 MΩ cm) from a water purification system (ELGA LabWater, Veolia Water Technologies, Celle, Germany) and sub-boiled nitric acid were used for the preparation of all solutions. The multi-element solution IV from Inorganic Ventures (Christiansburg, USA) was employed at a concentration of 50 ng g−1 for daily optimization of the performance of the MC-ICP-MS. In liquid mode mass bias was measured with a 10 ng g−1 solution of the isotopic reference material CRM U-500 (New Brunswick Laboratory, NBL, Argonne, IL, USA) isotopic reference material, while CRM U-020 (New Brunswick Laboratory, Argonne, IL, USA) was used to check the measured U isotope ratios. CRM solutions were prepared by dissolving an aliquot of the solid reference material in 8 M nitric acid with gentle heating. For LA analysis, a pellet of CRM U-020 was used to measure the mass bias factor and to calculate the gains of the employed ion counters. The accuracy of measurements was checked through repeated analysis of a pellet of CRM U-010 (New Brunswick Laboratory, Argonne, IL, USA). Further details on the measurement have been reported previously [20]. 2.3. Investigated samples The spatial distribution of the U isotopic composition was assessed with LA-MC-ICP-MS involving various solid U bearing materials, including certified reference materials (CRMs) and U samples, containing either depleted or low-enriched U: (a) a portion of the powdered CRM U-010 (NBL, Argonne, IL, USA) was pressed into a pellet with an XPRESS hydraulic laboratory press (Spex Industries, Metuchen, USA) applying a force of 2 tons for 8 min. This way, stable pellets were produced without using any kind of additional binder. This U3O8 pellet with a diameter of 5 mm and ~1 mm height was subsequently subjected to LAMC-ICP-MS. (b) CRM 125-A consisted of a sintered low-enriched (~4%) uranium oxide (UO2) pellet with a mass of ~5.4 g. The batch of pellets had been prepared by Westinghouse Commercial Nuclear Fuels Division (Columbia, SC, USA) and chemically characterized by NBL [5]. Both CRM U-010 and CRM 125-A are certified for their U isotope amount ratio together with the respective uncertainties at the 95% level of confidence. (c) two UO2 pellets from the 5th Collaborative Materials Exercise (CMX-5) organized by ITWG [23], and (d) two seized illicit U samples, here referred to Sample A and Sample B, respectively.

2. Experimental methods 2.4. Measurement procedures 2.1. Instrumentation All isotopic measurements were carried out with a NuPlasma™ (NU Instruments, Oxford, United Kingdom) double-focusing multi-collector inductively coupled plasma mass spectrometer (MC-ICP-MS). For U isotope ratio measurements the instrument was operated in low mass resolution mode (R = 300) with the major U isotopes (235U and 238U) measured on Faraday detectors, while ion counters were employed for the low abundant U isotopes (234U and 236U). As the n(235U)/n(238U) ratio is the most important parameter for the principal aim of this study, the main focus is set on this U isotope ratio rather than on the minor ones, i.e. n(234U)/n(238U) and n(236U)/n(238U), respectively. Sample introduction into the MC-ICP-MS was performed for aqueous samples via a T-connection through an Aridus II desolvating nebulizer (CETAC Technologies Inc., Omaha, NE, USA) or through a NWR213 laser ablation unit (ESI, Huntingdon, UK) in the case of solid samples. The ns-LA system was equipped with a two-volume cell (a socalled TV2 cell) largely helping to prevent cross-contamination from

After daily optimization of the MC-ICP-MS performance using aqueous calibration solutions, no more liquid samples were aspirated through the desolvating nebulizer. Instead small (~5 μm) spots were subsequently ablated with the laser from solid samples placed in the LA chamber. The T-connection between the desolvating nebulizer and the LA system allowed the immediate switching between liquid and solid sample introduction [20]. The particle cloud generated by LA was transported via a laminar argon gas stream to the inductively coupled plasma, where particles were atomized and ionized before entering the mass spectrometer. The U isotopic abundances of actual solid samples were determined by both single spot and line scan analysis. For single spot analysis, 5 μm or 8 μm spots were ablated at a frequency of 4 Hz for a period of 5 s. The applied fluence (laser power) was varied between 2 and 4 J cm−2 to adjust the 238U ICP-MS signal not to exceed 10 V that would otherwise cause saturation of the Faraday detector. Line scan analysis was performed ablating spots with a diameter of 5 μm. The focussed laser beam was moved across the surface of the


investigated solid samples along a pre-defined line with a speed of 20 μm s−1 at 4 Hz for several minutes. The fluence of the laser (2–4 J cm−2) was adjusted to generate a 238U signal of 4–8 V. No pre-ablation of the surface of any investigated sample was conducted. A pressed pellet of CRM U-020 was analyzed via single spot analysis between all sample measurements (either single spot or line scan analysis). All reported U isotope ratios were corrected for mass bias, measuring CRM U-020 as a reference standard prior to any sample.

0.0104

n(U-235)/n(U-238)

26

0.0103 0.0102 0.0101 0.0100 0.0098

n(U-235)/n(U-238)

3. Results and discussion Dissolving an aliquot of a specific solid sample will always generate averaged results for any measurement parameter. Consequently, sample inhomogeneity can be hardly studied this way. The high spatial resolution of LA (b1 μm to a few μm, depending on the applied laser system) has the potential for the fast and accurate investigation of microinhomogeneity within a U sample.

3.2. Homogeneity of UO2 pellet - CRM 125-A As mentioned above, the certified reference material CRM U-010 was pressed to a pellet and thus represents an idealized matrix of a

0

100

200

300

400

500

600

700

analysis number

0.0428 0.0426 0.0424 0.0422 0.0420

B

0.0418

3.1. Potential of LA-MC-ICP-MS for homogeneity studies

0

1000

2000

3000

4000

5000

6000

analysis number

frequency

Initial experiments, aiming at testing the capabilities of the LA-MCICP-MS set-up for homogeneity testing of solid U bearing materials with respect to their U isotopic composition, were performed with the certified U isotopic standard reference material CRM U-010 that is made of highly purified U3O8. This powdered CRM was pressed into 5mm pellets to allow easy and straightforward handling of the radioactive material and to avoid cross-contamination with other samples simultaneously present in the LA cell. Line scan analysis sampled LA spots of 5 μm diameter each across the surface of the pressed reference material yielding 695 data points (Fig. 1A). The experimentally determined n(235U)/n(238U) ratio of CRM U-010 revealed a very small spread among the data points with most of them being centred around the mean of the certified value (Fig. 1A). Therefore, the experimental n(235U)/n(238U) ratio of 0.01014 ± 0.00004 (relative standard deviation (RSD) is 0.40%) matched approvingly with the certified value of 0.01014 ± 0.00003 (Table 1). To further evaluate method performance, the absolute difference between the mean measured and certified n(235U)/n(238U) ratio was compared to the expanded uncertainty of the measured and certified n(235U)/n (238U) ratio. This statistical test indicated that there is no significant difference between the measurement results and the certified values of both CRMs. Additionally, the frequency histogram of the n(235U)/n (238U) ratio, consisting of almost 700 data points, displayed a narrow, symmetric Gaussian-like shape (Fig. 4A). A recent study reported n (235U)/n(238U) ratios measured in 16 individual particles of CRM U010 using high sensitivity LA-single collector sector field ICP-MS with an external precision of 4% [13]. The precision of LA-MC-ICP-MS performed on bulk U samples in this study is an order of magnitude better. All together, these results allow two important conclusions: first, the applied LA-MC-ICP-MS procedure generates both accurate and precise n (235U)/n(238U) ratio data of the pressed U reference material (Table 1). That implies that the small fluence applied for LA did not cause any measurable fractionation of the U isotopic measurements. Second, the small standard deviation of the mean of the experimental data set together with the narrow, symmetric distribution of the n(235U)/n(238U) ratio indicates that the investigated U material is homogeneous at the micrometer level. While U isotopic homogeneity of the material is to be expected because the reference material has been produced by hydrolyzing UF6, followed by crystallising it as U nitrate from an aqueous solution [24], it is advantageous to demonstrate that the developed analytical technique can convincingly reproduce this fact.

A

0.0099

1400 1200 1000 800 600 400 200 0

C 0.0420

0.0422

0.0424

0.0426

n(U-235)/n(U-238) Fig. 1. n(235U)/n(238U) ratio analysis using LA-MC-ICP-MS. (A) Line scan analysis of a pressed pellet of CRM U-010 exemplified by 695 data points. The red solid line represents the certified U isotopic ratio, while the dotted lines indicate the corresponding uncertainty. (B) Line scan analysis of CRM 125-A, a UO2 pellet employing 5703 data points underpinning the good agreement between the certified and experimental n(235U)/n(238U) ratios. (C) Frequency histogram of the n(235U)/n(238U) ratios determined in CRM 125-A. The narrow, symmetric, Gaussian-like distribution of the n(235U)/n(238U) ratios indicates that the U isotopic composition is uniform in the investigated UO2 pellet. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

homogeneous U material. To further test the applicability of the above analytical procedure, additional line scan experiments were conducted on a solid UO2 pellet (CRM 125-A). More than 5700 individual data points were acquired throughout the measurement of this CRM 125-A isotopic standard reference material (Fig. 1B). Again, all experimental n(235U)/n(238U) ratios were centred around the certified value and its corresponding reported uncertainty (Fig. 1B). The measured n(235U)/n(238U) ratio of 0.042290 ± 0.000088 (RSD 0.21%) agreed well with the certified value of 0.042301 ± 0.000025 underpinning both the accuracy and precision of the applied analytical procedure (Table 1). It is important to note that the precision achieved here for this CRM is about an order of magnitude

Table 1 Experimental (average ± standard deviation) LA-MC-ICP-MS and certified n(235U)/n (238U) ratios of two certified reference materials as determined via line scan analysis of 5 μm spots. na

Measured

Certified

CRM U-010 Pressed powder

695

0.01014 ± 0.00004

0.01014 ± 0.00003

CRM 125-A UO2 pellet

5703

0.042290 ± 0.000088

0.042301 ± 0.000025

a

Number of acquired data points during the measurement sequence.


Prior to LA-MC-ICP-MS analysis, several fractions of each of the two UO2 pellets from the 5th CMX exercise were dissolved in 8 M nitric acid. These concentrated solutions were further diluted with high purity water for subsequent solution-based U isotopic analysis using TIMS and MC-ICP-MS. Consistent results were obtained only when the identical solution of one sub-sample was analyzed for its U isotopic composition for one of the pellets (data not shown). However, when different fractions of a pellet were dissolved and subsequently analyzed for their U isotopic composition, results between the used techniques did not match (data not shown). This observed discrepancy suggested that the investigated UO2 pellets are inhomogeneous with respect to their U isotopic composition. To shed further light into this aspect, LA-MCICP-MS analysis was applied to the solid UO2 pellets to investigate their U abundance at the microscopic level. 3.3.1. Individual laser ablation spot analysis Initial LA experiments focussed on the U isotopic analysis of individual spots having a diameter of 5 μm. After completion of one analysis, the laser beam was moved for 30 μm to conduct the subsequent measurement. This procedure was followed along a horizontal line for some more LA spots covering a distance of 0.33 mm (11 points for CMX-5/1) and 0.45 mm (15 points for CMX-5/2), respectively. For the first UO2 pellet, named hereafter CMX-5/1, the obtained n (235U)/n(238U) ratio varied between ~0.008 and ~0.025 for 11 individual LA spots, indicating that the material is inhomogeneous with respect to its U isotopic composition at this size level. This large variation of the U abundance was even more pronounced for the second UO2 pellet, i.e. CMX-5/2 (Fig. 2). The n(235U)/n(238U) ratio for 15 individual LA spots ranged from ~0.006 to 0.038, highlighting the even higher variability of the 235U enrichment of CMX-5/2 compared to that of CMX-5/1. 3.3.2. Line scan analysis In order to study the range of the n(235U)/n(238U) ratio of both UO2 pellets in more detail, over 4100 individual analyzes each were carried out for both CMX-5/1 and CMX-5/2 employing a line scan. As such, the 5 μm laser beam, operated at a frequency of 4 Hz, was moved with 0.04

CMX-5/1 CMX-5/2

n(U-235)/n(U-238)

3.3. Intra-pellet homogeneity of two UO2 pellets from the 5th CMX

a speed of 20 μm s−1 across the surface of the pellet. These line scans covered a representative distance of N2 cm of the surface of each of the UO2 pellets (Fig. 3A + B). After evaluating this experiment for CMX-5/1, it was obvious that the distribution of the U isotopic composition within the pellet is not uniform. Fig. 3A highlights several distinct areas within the surface of the UO2 pellet that are clearly enriched in 235U relative to the average that is centred around ~1% 235U enrichment. While the average n (235U)/n(238U) ratio was 0.00963 ± 0.00094 (RSD 9.7%), the large spread ranging from 0.00755 to 0.01589 clearly revealed the inhomogeneity of the investigated CMX-5/1 material. As already observed during the individual spot analysis of sample CMX-5/2 (Fig. 2), line scan analysis confirmed the even larger heterogeneity of this UO2 pellet (Fig. 3B). The n(235U)/n(238U) ratio essentially ranged from 0.00520 to 0.02975, with an average of 0.01142 ± 0.00403 (RSD 35%). The ~3.5-times larger RSD associated with the average n(235U)/n(238U) ratio of CMX-5/2 was another proof for the increased heterogeneity of CMX-5/2 compared to that of CMX-5/1. Again distinct areas of 235U enrichments below and exceeding the average of ~1.1% were found within this pellet. It is important to note that these areas of higher and lower 235U enrichment are not artefacts of the measurement, but are frequently supported by N50 consecutive data points. In other words, these values are no outliers due to a poor measurement, but represent actual areas of varying 235U enrichment. To put the heterogeneity of the n(235U)/n(238U) ratio of CMX-5/1 and CMX-5/2 into perspective, Fig. 3C displays the line scan analyzes of both CMX-5 samples together with that of CRM U-010 on the same y-axis, providing the identical scale of the n(235U)/n(238U) ratio for all three samples. In this context, the results for CRM U-010 are

n(U-235)/n(U-238)

superior to that reported for LA-single collector sector field ICP-MS [5]. Similarly to CRM U-010, the frequency histogram of the established n (235U)/n(238U) ratios of CRM 125-A yielded a narrow and symmetric, Gaussian-like distribution (Fig. 1C), stressing the homogeneity of the UO2 pellet at the μm-level.

0.016

A

0.014 0.012 0.010 0.008 0.006

0

n(U-235)/n(U-238)

n(U-235)/n(U-238)

0.01

1000

2000

3000

4000

0.030

B

0.025 0.020 0.015 0.010 0.005 0

0.03

0.02

27

1000

2000

3000

4000

CMX-5/2 CMX-5/1 U-010

0.030 0.025 0.020

5000

C

0.015 0.010

offset -0.05

0.005 0

1000

2000

3000

4000

analysis number 0

3

6

9

12

15

analysis number Fig. 2. Single spot LA-MC-ICP-MS analysis of n(235U)/n(238U) ratios of two UO2 pellets from the 5th Collaborative Materials Exercise (CMX-5) highlighting the high variability of the n(235U)/n(238U) ratios within each pellet.

Fig. 3. Line scan LA-MC-ICP-MS analysis of n(235U)/n(238U) ratios of two UO2 pellets from the 5th Collaborative Materials Exercise (CMX-5): (A) more homogeneous pellet CMX-5/1 with n(235U)/n(238U) ratios ranging from 0.00755 to 0.01589, (B) less homogenous pellet CMX-5/2 with n(235U)/n(238U) ratios ranging from 0.00520 to 0.02975, and (C) direct comparison of CMX-5/1, CMX-5/2, and CRM U-010. Note that an offset of −0.05 was applied to the actual values of CRM U-010 to increase the readability of the figure.

28

M. Krachler et al. / Microchemical Journal 140 (2018) 24–30 100

CRM U-010

frequency

80

80

60

60

40

40

20

0 0.0100

20

0.0101

0.0102

0.0103

A 0

0.01

0.02

0.03

n(U-235)/n(U-238)

2000

CMX-5/1

frequency

1500

1000

500

B 0

0.01

0.02

0.03

n(U-235)/n(U-238)

500

frequency

400

applied LA-MC-ICP-MS methodology for homogeneity studies with respect to the U isotopic composition of solid U bearing materials. Because N4100 n(235U)/n(238U) ratios were measured throughout the line scans of each CMX-5 sample, meaningful frequency histograms could be generated (Fig. 4). This data provides an additional useful indicator for inhomogeneity of U isotopic abundance and shows the mixed components used for U production. The asymmetric frequency histograms of the distribution of the n(235U)/n(238U) ratio for both CMX-5/ 1 (Fig. 4B) and CMX-5/2 (Fig. 4C) are quite different compared to those established for the two CRMs U-010 (Fig. 4A) and CRM 125-A (Fig. 1C). The related histograms of the two CMX-5 samples reveal a distinct tailing towards larger n(235U)/n(238U) ratios (Fig. 4B and C) reflecting the inhomogeneity of the U isotopic composition of each of the two UO2 pellets. While the majority of the 235U enrichments for CMX-5/1 ranged from 0.92–0.96% (total range: 0.75%–1.6%), the corresponding bulk of data points for CMX-5/2 spanned a much broader range of 0.7–1.4% (total range: 0.45%–3.0%). Interestingly, both UO2 pellets distributed to the CMX-5 participants were prepared by the French Alternative Energies and Atomic Energy Commission (CEA, Cadarache, France) from the same batches of starting materials [25]. While 0.3% depleted 235U and 4.3% enriched 235U were used at identical proportions for the production of both UO2 pellets, their manufacturing process was different. Normal Double Cycle (NDC) with compaction of the employed U fractions at low pressure (50 MPa) was applied to CMX-5/1, whereas CMX-5/2 was prepared using Inverse Double Cycle (IDC) at high pressure (600 MPa) mixing the components later [25]. Employing both production routes, the target value of 1% 235U enrichment was achieved satisfactorily for both UO2 pellets. However, according to CEA internal isotopic analysis, CMX-5/1 was more homogeneous with a 235U enrichment of 1.0063 ± 0.0009 (n = 3, RSD 0.09%) than CMX-5/2 that was slightly higher enriched in 235U, i.e. 1.0158 ± 0.0232 (n = 3, RSD 2.29%) [25]. These values are in excellent agreement with our findings. Such real world samples manufactured from the same starting materials, but being produced by different processes reveal different heterogeneity of the final product. In turn, in a nuclear forensic investigation the isotopic homogeneity of industrial material (e.g. nuclear fuel pellets) can be exploited to provide hints on the production process. LAMC-ICP-MS is a powerful and straightforward analytical tool to

CMX-5/2

300

200

100

0

C 0.01

0.02

0.03

n(U-235)/n(U-238) Fig. 4. Frequency histograms of n(235U)/n(238U) ratios of (A) reference material CRM U010 as well as of the two UO2 pellets CMX-5/1 (B) and CMX-5/2 (C) from the 5th Collaborative Materials Exercise (CMX-5). Using the identical x-axis scale for all three plots, the homogeneity of CRM U-010 (A) and the dissimilar inhomogeneity of CMX-5/1 and CMX-5/2 with respect to their n(235U)/n(238U) ratios are evident. The inserted figure in (A) shows an expanded view of the n(235U)/n(238U) ratios demonstrating the narrow frequency distribution in more detail.

represented by a flat horizontal line indicating a homogenous U isotopic composition of this material. For a better readability of the figure, an offset of −0.05 was applied to the actual values of the n(235U)/n(238U) ratio for CRM U-010. Heterogeneity of the n(235U)/n(238U) ratio is clearly evident for CMX-5/1 and is even more pronounced for CMX-5/ 2. Taken together, Fig. 3C favourably highlights the potential of the

Fig. 5. Optical microscopy image of yellow and black particles present in seized Sample B, showcasing the potential heterogeneity of the material. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)


29

Table 2 Uranium isotope ratios of black and yellow areas of two seized U samples (Sample A and Sample B) as determined on several individual 5 μm spots using LA-MC-ICP-MS (average ± stanndard deviation).

Sample A Black area Yellow area Sample B Black area Yellow area a

na

n(234U)/n(238U)

n(235U)/n(238U)

n(236U)/n(238U)

9 9

0.0000219 ± 0.0000003 0.0000218 ± 0.0000003

0.003722 ± 0.000008 0.003733 ± 0.000015

b2 × 10−6 b2 × 10−6

6 6

0.0000246 ± 0.0000020 0.0000266 ± 0.0000004

0.004021 ± 0.000200 0.004166 ± 0.000039

b2 × 10−6 b2 × 10−6

Number of individual 5 μm spots analyzed.

investigate isotopic (in-)homogeneity at the micrometer scale with good accuracy and precision. 3.4. Seized U-bearing materials The developed analytical methodology was also applied to two seized U metal samples that had been found out of regulatory control. Initially, fractions of the two samples were dissolved in nitric acid, followed by solution-based analysis of their U isotopic composition using TIMS and MC-ICP-MS. The obtained results agreed well (data not shown) between the two instrumental techniques and revealed the following information: (a) both specimens contain depleted U, but (b) the 235U abundance is significantly different in the two samples, and (c) the same holds true for the abundances of 234U and 236U. From a nuclear forensics perspective all this data pointed to either different source materials or different amounts of the source materials in the two seized samples. Both samples, hereafter named Sample A and Sample B, consisted of macroscopically-evident black and yellow areas (Fig. 5), the U isotopic analysis of which might provide another hint regarding the source material. Therefore, these differently coloured areas of the two solid U samples were further investigated for their U isotopic abundance using LAMC-ICP-MS (Table 2). For Sample A, there was no difference in any of the U isotope ratios between the black and yellow areas (Table 2). The established n(235U)/n(238U) ratios in the solid Sample A matched well with the corresponding value of 0.003731 ± 0.000015 (average ± expanded uncertainty, k = 2) obtained for the dissolved sample. The black areas of Sample B, however, revealed increased inhomogeneity of the n(234U)/n(238U) and n(235U)/n(238U) ratios as reflected by higher standard deviations of their mean values compared to the other LA-MCICP-MS measurements (Table 2). Again, the n(235U)/n(238U) ratio in the solid was consistent with the corresponding solution-based value of 0.004216 ± 0.000013. Taken together, no difference in the U isotopic composition within uncertainties between the different regions of the two samples could be established suggesting that both materials are homogeneous at the LA measurement level. However, black areas of Sample B revealed larger variability in their U isotopic composition, therefore inhomogeneity cannot be excluded. However, the general good match between the U isotopic composition between black and yellow regions as well as the solution-based bulk measurement for each of the two seized U samples indicates that both materials are homogeneous. Consequently, the colour differences observed for both investigated samples might arise from different degrees of oxidation of the current U (yellow regions representing U at oxidation state +6) and are not associated with isotopic inhomogeneity within the material. As the 235U/238U ratio differs significantly between Sample A and Sample B (Table 2), both samples have been prepared from different source materials or using different amounts of them. 4. Conclusions LA-MC-ICP-MS proofed to be a fast and powerful tool to investigate the (in-)homogeneity of the U isotopic composition of solid U-bearing

materials. The minute amount of material required for analysis renders LA-MC-ICP-MS especially suited for nuclear forensics applications, as the analyzed sample (evidence) is readily available and only altered marginally for following investigations. Repetitive U isotopic analysis of a pressed nuclear reference material (CRM U-010), a certified UO2 pellet (CRM 125-A), two UO2 pellets from the CMX-5 exercise as well as two seized illicit U samples yielded 235U/238U ratios that were in excellent agreement with either the certified values (CRM U-010 and CRM 125-A) or solution-based analysis using TIMS or MC-ICP-MS. These features demonstrate LA-MC-ICP-MS an excellent analytical approach for nuclear forensics delivering accurate results within a short time frame. Future studies will extend the application of LA-(MC)-ICP-MS to the determination of elemental impurities in nuclear materials, U isotope ratio measurements of particles and isotopic measurements of several impurities (Pb, Sr or Nd) for origin assessment. In addition, the determination of the production age of U materials - based on the analysis of the 230 Th/234U ratio - is envisaged. If an unknown U material consists of two or more different components, these could have two or more production dates. Employing LA-MC-ICP-MS, this information is accessible providing additional valuable information to law enforcement investigations to trace seized nuclear material that got out of regulatory control back to its origin.

Acknowledgements Comparative U isotopic analysis of the two CMX-5 samples and the two seized U samples employing TIMS by the Analytical Service of JRC Karlsruhe is gratefully acknowledged. M. Ernstberger, JRC Karlsruhe, provided the optical microscopy picture of the seized U sample. References [1] M.J. Kristo, A.M. Gaffney, N. Marks, K. Knight, W.S. Cassata, I.D. Hutcheon, Nuclear forensic science: analysis of nuclear material out of regulatory control, Annu. Rev. Earth Planet. Sci. 44 (2016) 555–579. [2] 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. [3] Z. Varga, K. Mayer, C.E. Bonamici, A. Hubert, I. Hutcheon, W. Kinman, M. Kristo, F. Pointurier, K. Spencer, F. Stanley, R. Steiner, L. Tandon, R. Williams, Validation of reference materials for uranium radiochronometry in the frame of nuclear forensic investigations, Appl. Radiat. Isot. 102 (2015) 81–86. [4] M. Krachler, R. Alvarez-Sarandes, G. Rasmussen, High resolution inductively coupled plasma optical emission spectrometry for 234U/238Pu age dating of plutonium materials and comparison to sector field inductively coupled mass spectrometry, Anal. Chem. 88 (2016) 8862–8869. [5] R.C. Marin, J.E.S. Sarkis, M.R.L. Nascimento, The use of LA-SF-ICP-MS for nuclear forensics purposes: uranium isotope ratio analysis, J. Radioanal. Nucl. Chem. 295 (2013) 99–104. [6] T.L. Spano, A. Simonetti, E. Balboni, C. Dorais, P.C. Burns, Trace element and U isotope analysis of uraninite and ore concentrate: applications for nuclear forensic investigations, Appl. Geochem. 84 (2017) 277–285. [7] S. Bürger, L.R. Riciputi, A rapid isotope ratio analysis protocol for nuclear solid materials using nano-second laser-ablation time-of-flight ICP-MS, J. Environ. Radioact. 100 (2009) 970–976. [8] K. Ito, N. Hasebe, A. Hasebe, S. Arai, The matrix effect on 238U and 232Th measurements using pressed powder pellets by LA-ICP-MS, Geochem. J. 45 (2011) 375–385. [9] S.J. Walsh, N. Dzigal, E. Chinea-Cano, A. Limbeck, Simple robust estimation of uranium isotope ratios in individual particles from LA-ICP-MS measurements, J. Anal. Atom. Spec. 32 (2017) 1155–1165.

30


[10] Z. Varga, Application of laser ablation inductively coupled plasma mass spectrometry for the isotopic analysis of single uranium particles, Anal. Chim. Acta 625 (2008) 1–7. [11] F. Pointurier, A. Hubert, A.-C. Pottin, Performance of laser ablation: quadrupolebased ICP-MS coupling for the analysis of single micrometric uranium particles, J. Radioanal. Nucl. Chem. 296 (2013) 609–616. [12] A. Hubert, F. Claverie, C. Pécheyran, F. Pointurier, Measurement of the isotopic composition of uranium micrometer-size particles by femtosecond laser ablationinductively coupled plasma mass spectrometry, Spectrochim. Acta B 93 (2014) 52–60. [13] A. Donard, F. Pointurier, A.-C. Pottin, A. Hubert, C. Pécheyran, Determination of the isotopic composition of micrometric uranium particles by UV femtosecond laser ablation coupled with sector-field single-collector ICP-MS, J. Anal. Atom. Spec. 32 (2017) 96–106. [14] Z. Stefánka, R. Katona, Z. Varga, Laser ablation assisted ICP-MS as a tool for rapid categorization of seized uranium oxide materials based on isotopic composition determination, J. Anal. Atom. Spec. 23 (2008) 1030–1033. [15] A. Kuchkin, V. Stebelkov, K. Zhizhin, Ch. Lierse von Gostomski, Ch. Kardinal, A.H.J. Tan, B.K. Pong, E. Loi, E. Keegan, M.J. Kristo, M. Totland, I. Dimayuga, M. Wallenius, Contribution of bulk mass spectrometry isotopic analysis to characterization of materials in the framework of CMX-4, J. Radioanal. Nucl. Chem. (2018)https://doi.org/ 10.1007/s10967-017-5681-4. [16] S.F. Boulyga, T. Prohaska, Determining the isotopic compositions of uranium and fission products in radioactive environmental microsamples using laser ablation ICPMS with multiple ion counters, Anal. Bioanal. Chem. 390 (2008) 531–539. [17] N.S. Lloyd, R.R. Parrish, M.S.A. Horstwood, S.R.N. Chenery, Precise and accurate isotopic analysis of microscopic uranium-oxide grains using LA-MC-ICP-MS, J. Anal. Atom. Spec. 24 (2009) 752–758.

[18] S. Kappel, S.F. Boulyga, T. Prohaska, Direct uranium isotope ratio analysis of single micrometer-sized glass particles, J. Environ. Radioact. 113 (2012) 8–15. [19] W. Wang, Z.-M. Li, J. Xu, G.-Q. Zhou, F. Wang, X.-P. Shen, L.-H. Zhai, C.-F. Ke, Determination of 235U/238U ratio in uranium particles by laser ablation multi-collector inductively coupled plasma mass spectrometry, J. Chin. Mass Spectrom. Soc. 37 (2016) 213–221. [20] Z. Varga, M. Krachler, A. Nicholl, M. Ernstberger, T. Wiss, M. Wallenius, K. Mayer, Accurate measurement of uranium isotope ratios in solid materials by laser ablation multi-collector inductively coupled plasma mass spectrometry, J. Anal. Atom. Spec. (2018) (in press). [21] V. Stebelkov, I. Elantyev, M. Hedberg, M. Wallenius, A.-L. Fauré, Determination of isotopic composition of uranium in the CMX-4 samples by SIMS, J. Radioanal. Nucl. Chem. 315 (2018) 347–352. [22] ITWG Nuclear Forensics Update, No. 1, http://www.nf-itwg.org/newsletters/ITWG_ Update_no_1.pdf December 2016 (accessed 01 February 2018). [23] ITWG Nuclear Forensics Update, No. 4, http://www.nf-itwg.org/newsletters/ITWG_ Update_no_4.pdf September 2017 (accessed 01 February 2018). [24] E.L. Garner, L.A. Machlan, W.R. Shields, Standard Reference Materials: Uranium Isotopic Standard Reference Materials (Certification of Uranium Isotopic Standard Reference Materials), National Bureau of Standards Special Publication 260-27, April 1971 162 pages. [25] P. Girard, Manufacturing & characterizations of two batches of UO2 pellets for ITWG/ CMX-5, Presentation Given at the CMX-5 Data Review Meeting, Bucharest, Romania, April 26–28, 2017.