Microscale Isotopic Variation in Uranium Fuel Pellets with Implications

Subscriber access provided by GUILFORD COLLEGE

Article

Microscale Isotopic Variation observed in Uranium Fuel Pellets with implications for Nuclear Forensics Ruth Kips, Peter K. Weber, Michael J. Kristo, Benjamin Jacobsen, and Erick Ramon Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.9b01737 • Publication Date (Web): 16 Jul 2019 Downloaded from pubs.acs.org on July 23, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Microscale Isotopic Variation observed in Uranium Fuel Pellets with implications for Nuclear Forensics Ruth Kips*, Peter K. Weber*, Michael J. Kristo, Benjamin Jacobsen, Erick Ramon Lawrence Livermore National Laboratory, P.O. Box 808, Livermore CA 94551, USA *Correspondence: [email protected]; [email protected]

Abstract Until recently, the analysis and identification of nuclear fuel pellets in the context of a nuclear forensics investigation have been mainly focused on macroscopic characteristics such as fuel pellet dimensions, uranium enrichment, and other reactor-specific features. Here we report microscale isotopic heterogeneity observed in different fuel pellet fragments that were characterized in situ by NanoSIMS. The materials analyzed include fuel fragments obtained as part of the Collaborative Materials Exercise (CMX-4) organized by the Nuclear Forensics International Technical Working Group (ITWG), as well as a fuel pellet fragment from a commercial power reactor. While the commercial fuel pellet showed a homogeneous 235U/238U ratio across the sample (within analytical error), NanoSIMS imaging of the CMX-4 fuel pellet fragments showed distinct microscale variations in the uranium isotopic composition. The average 235U enrichment was 2.2 and 2.9 % for the two samples respectively, however, the measured 235U/238U ratios varied between 0.0081 - 0.035 (0.79 – 3.3 at. % 235U) and 0.0090 - 0.045 (0.89 - 4.3 at. % 235U). The measurement of the 236U in one of the CMX-4 samples suggested the presence of at least three uranium oxide powders of different isotopic composition (‘source terms’) used in the production of the pellets. These variations were not detected using the conventional bulk/macroscopic techniques applied to these materials. Our study highlights the importance of characterizing samples on the microscale for heterogeneities that would have otherwise been overlooked and demonstrates its potential use in guiding further nuclear forensic analysis.

1 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 18

Introduction Uranium dioxide is the most commonly used fuel type in nuclear reactors worldwide, with fuel fabrication facilities producing hundreds of uranium pellets per minute.1 Illicit trafficking of uranium dioxide fuel pellets is, in fact, a reoccurring phenomenon.2-4 The majority of confirmed cases of illicit trafficking of nuclear materials that were reported to the IAEA Incident and Trafficking Database (ITDB) involved low-grade nuclear materials, i.e. natural uranium, depleted uranium and low enriched uranium (LEU), often in the form of reactor fuel pellets.2 These cases are indicative of gaps in the control and security of certain nuclear material and nuclear facilities. 2 Fortunately, each fuel manufacturer applies a somewhat different set of technological processes to the material, which can help trace the material back to the originating fuel fabrication facility.5 Until recently, however, the analysis of fuel pellets in the context of a nuclear forensics investigation has mostly focused on sample characteristics such as fuel pellet dimensions, bulk concentrations of impurities, and uranium isotopic abundances.6 Investigating the uranium isotopic variability on the microscale was triggered by the results from the ITWG CMX-4 sample comparison exercise, where three of the participating laboratories performed isotopic analysis on particles associated with the fuel pellet samples using secondary ion mass spectrometry (SIMS).7, 8 These particles were collected either by swiping the particles from the surface of the pellet samples and/or the packaging material in which the samples were received. Bulk mass spectrometric analyses determined the two CMX-4 pellet samples to have a 235U abundance of 2.2 % and 2.9 % respectively, whereas the SIMS measurements on the particles associated with the samples showed a range of isotopic compositions, with uranium enrichments of up to 4.5 %.7 To determine whether this variation in enrichment was inherent to the samples, nanoscale secondary ion mass spectrometry (NanoSIMS) was used at LLNL to probe for uranium isotopic heterogeneity directly in the pellet. Scanning electron microscopy (SEM) images were collected before and after the NanoSIMS analyses to correlate the observations in the ion images (uranium isotope ratios) with the fuel pellet surface morphology. In addition to the CMX-4 fuel pellet fragments, the uranium isotopic ratio distribution of a light water graphite-moderated RBMK pellet fragment.10, 11 was analyzed for comparison.

Experimental Three LEU, unirradiated, solid fragments from crushed fuel pellets were used for this study. Two pellet samples obtained during the CMX-4 exercise (ES-2 and ES-3) had an average 235U/238U ratio of 0.0225 (2.2 at. % 235U) and 0.0298 (2.9 at. % 235U) respectively, as measured by inductivelycoupled plasma mass spectrometry (ICP-MS).7, 9 Sample ES-3 had a detectable isotopic abundance of 236U (~2×10–5); sample ES-2 had no detectable concentration of 236U (≤1×10–7).9 The RBMK pellet fragment (RBMK-1) had an average 235U/238U ratio of 0.0207 (2.1 at. % 235U).10, 11 Each of the fragments was embedded in epoxy in millimeter-sized stainless-steel bullets for analysis by optical and electron microscopy, before and after nanoscale secondary ion mass spectrometry (NanoSIMS). The surface of the pellet samples was exposed and flattened with a silicon carbide grit series and then polished using a diamond paper series (45 µm, 30 µm, 15 µm, 9 µm, 6 µm, 3 µm, 1 µm) and finished with 1 µm diamond paste. The CMX-4 samples (ES-2 and ES-3) were


Page 3 of 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


initially analyzed without applying a coating, but all three samples were then coated with either a thin gold or carbon layer to minimize charging.

Optical Microscopy A Nikon Optiphot2-Pol Light Microscope with Olympus DP72 camera was used for the initial characterization of the fuel pellet fragments. This instrument has an excellent depth of field, can be used in reflected, transmitted or polarizing light mode and produces color images of the entire pellet fragments. Although the image magnification for optical microscopy is not as high as for electron microscopy, the areas analyzed by NanoSIMS were often more easily relocated using optical microscopy, especially when the samples are uncoated.

Electron Microscopy A high brightness Schottky field emission gun (FEG) SEM (Inspect F, FEI) with an EverhartThornley secondary electron detector, a solid-state diode backscatter electron detector and an energy-dispersive X-ray detector (QUANTAX, Bruker) was used to perform an initial characterization of both the morphology and elemental composition of the uranium fuel pellet samples.

NanoSIMS High-spatial resolution ion images were obtained with the LLNL NanoSIMS 50 (Cameca, Gennevilliers, France). This instrument is outfitted with the prototype Hyperion II radio frequency inductively coupled plasma ion source (Oregon Physics, Hillsboro, Oregon, USA). A 16 keV beam of O- primary ions was rastered over the sample surface, and positive secondary ions were extracted into a double focusing mass spectrometer tuned for ~3000 mass resolving power (M/M; 1.5x correction).12 The ion species [234U+], [235U+, 235UO+], [236U+], [238U+, 238UO+], [238UH+] were monitored on electron multipliers (EMs) in pulse counting mode using magnetic field switching (a.k.a., peak jumping). Data were collected in two rounds: first to survey 235U/238U heterogeneity over larger areas (~20,000 µm2; 100 pA O- analysis beam, no 236U) and second to collect higher precision minor isotope data (~5,000 µm2; 500 pA O- analysis beam, 236U). The larger survey areas were analyzed with automated analysis sequences with rasters of 20×20 µm2 or 25×25 µm2. These rasters were separated by a few micrometers to facilitate individual raster relocation. After U isotopic heterogeneity was identified, we added 236U+ to later analyses and increased the primary ion beam current to collect sufficient counts. These rasters were 40×40 µm2. The counts on 238U1H+ were collected to correct for the contribution of 235U1H+ on m/z 236. The certified standards used for these analyses were the U3O8 particle standard CRM U200 (New Brunswick National Laboratory, 20.013 at. % 235U) and the UO2 fuel pellet CRM 125-A (New Brunswick National Laboratory, 4.0569 at. % 235U). Mass fractionation varied between 3% and 0% correction for the 235U/238U ratio between sessions and was corrected where applicable. The 235U/238U results based on U+ and UO+ were indistinguishable, and the 235U/238U data reported below are all based on the atomic U+ counts.



Page 4 of 18

Data visualization and extraction was performed using quantitative image processing software (L’IMAGE, Larry Nittler, Carnegie Institution of Washington, Washington, DC, USA, run in IDL, Harris Geospatial Solutions, www.harris.com). A deadtime correction of 44 ns was applied to the EMs. Ion ratio images for the uranium isotopes were generated by taking the ratio of the respective ions for each pixel and smoothing the data using a 7×7-pixel (~600-700 nm) boxcar routine. This smoothing routine was applied to avoid false heterogeneity in the image since the primary ion beam was on the order of 700 nm and considerably larger than one pixel. Numerical data were extracted for the analyzed areas by defining regions of interest (ROIs) automatically using a gridding routine (~4×4 µm2 ROIs) and manually (~3×3 µm2 ROIs) based on the isotope ratio images. Uncertainties are reported as 2 standard deviations (2SD). The uncertainties in the corrected 236U/238U data were calculated using Gaussian error propagation. Regressions were defined using Microsoft Excel.


Page 5 of 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Results Optical and SEM Imaging Optical and SEM images were collected before and after NanoSIMS analysis. The fuel pellet fragments analyzed in this study were approximately 5 mm wide, with sharp, chipped edges created during sub-sampling of the pellets. Polishing exposed voids, cracks, pore structures and other pellet internal features. Optical and electron microscopy showed that pore features were ubiquitous in all three samples analyzed but were significantly more abundant in the two CMX-4 samples where they formed larger networks of voids (Fig. 2). UO2 crystal boundaries (“grain boundaries”) were partially visible in the RBMK-1 sample but were not visible in the SEM images of the CMX-4 samples.

Fig 1. Optical microscope images of the respective fuel pellet samples embedded in epoxy in a stainless-steel bullet: Left: CMX-4 ES-2 (carbon-coated), middle: CMX-4 ES-3 (prior to carbon coating) and right: RBMK-1 (gold-coated).

Fig 2. Scanning electron microscope (SEM) images of polished sample ES-2 (left), ES-3 (middle) and RBMK-1 (right) showing voids, cracks, pore structures and other pellet internal features. Grain boundaries were partially visible in the RBMK-1 sample but were not visible in the SEM images of the CMX-4 samples ES-2 and ES-3.

NanoSIMS uranium isotopic composition analysis The two CMX-4 samples (ES-2 and ES-3) and the RBMK-1 pellet fragment were analyzed for uranium isotopic composition by NanoSIMS. Initially, the samples were mapped over ~20,000 µm2 areas with an emphasis on the 235U/238U composition. These areas were assessed for uranium



Page 6 of 18

isotopic homogeneity based on false-color 235U/238U maps (Fig. 3). Here, the color scale for 235U/238U is presented with the same range for all three samples because they all have a similar average 235U/238U composition. While the RBMK-1 sample showed a homogeneous 235U/238U ratio across the area analyzed (within analytical error), the CMX-4 ratio images suggested the 235U/238U ratios vary over nearly an order of magnitude. For the CMX-4 samples, areas with significantly higher or lower than the average ratio were up to several microns in size and were distributed across the pellet surface. Moreover, in some cases the areas with the highest ratios were separated by only a few micrometers from the areas with the lowest ratios (Fig. 3). The 234U/238U ratio images show a similar pattern, but with more noise because of the low precision data (not shown). Comparison of the 235U/238U ratio images with the 238U+ ion images and SEM images showed no direct correlation between these microscale uranium isotopic variations and the pellet grains (Fig. 3 & 4). The SEM analyses also indicated the CMX-4 samples ES-2 and ES-3 have more pores and areas of finer, irregularly shaped grains than the RBMK-1 pellet (Fig. 2 & 3), however, these microstructures were not directly related to isotopic heterogeneity. Although pores were ubiquitous across the analyzed surface, they did not affect the uranium isotope ratio analyses.


Page 7 of 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3: Subset of 235U/238U ratio images of sample ES-2, ES-3 and RBMK-1 and their corresponding 238U+ images. The color scale for 235U/238U is presented with the same range for all three samples. The ratio images (left) showed significant variation in the enrichment across the pellet surface for the CMX-4 samples compared to the RBMK-1 sample. Dark blue spots on these ion ratio images indicate areas of lower than average enrichment, while bright yellow spots represent areas with an enrichment higher than average. The 238U+ ion images (right) show the fuel pellets grains. There was no direct correlation observed between the areas of isotopic heterogeneity and the grains.



Page 8 of 18

Figure 4: Post-NanoSIMS SEM image (left) and corresponding NanoSIMS 238U+ ion image (middle) and 235U/238U ratio image (right) of a single 25×25 µm2 raster area on sample ES-2 (outside automated analysis sequence shown in Fig. 3). The color bar on the ratio image provides the isotope ratio scale. Although the grains are visible in both the SEM image and the 238U+ ion image, no direct correlation was observed between the areas of isotopic heterogeneity and the pellet grains or surface topography. The lines across the surface resulted from mechanical polishing.

Using all the automatically generated grid-based ROIs (~800 ROIs for each sample), the distribution of 235U/238U can be visually compared in a histogram (Fig. 5). The 235U/238U measurements for ES-2 and ES-3 range between 0.0081 - 0.035 (0.79 – 3.4 at. % 235U) and 0.0090 - 0.043 (0.89 - 4.1 at. % 235U), respectively. The variation in the 235U/238U ratios for the two CMX-4 samples (ES-2 and ES-3) is significantly larger than the RBMK-1 sample (standard deviation = 3.0×10-3, 4.8×10-3 and 1.3×10-4, respectively).

Figure 5: Histogram of the distribution of the 235U/238U ratios for the three samples, based on the automatically generated grid-based ROIs for all analyses.


Page 9 of 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


For the two CMX-4 samples ES-2 and ES-3, uranium isotope data and the respective three-isotope plots are shown in Fig. 6 and Fig. 7. Measurement precision (σ) for the grid-based ROIs (ES2 auto and ES3 auto) was ~0.5% for 235U/238U, 0.5 - 2% for 234U/238U, and ~5% for 236U/238U. The 236U was only analyzed for ES-3 because the bulk analyses indicated the 236U isotope was below detection for ES-2 (

Microscale Isotopic Variation in Uranium Fuel Pellets with Implications

Recommend Documents