Journal of Radioanalytical and Nuclear Chemistry https://doi.org/10.1007/s10967-018-5798-0 (0123456789().,-volV)(0123456789().,-volV)

A simple correction method for isobaric interferences induced by lead during uranium isotope analysis using secondary ion mass spectrometry Jinkyu Park1

•

Tae Hee Kim1 • Chi-Gyu Lee1 • Sang Ho Lim1,2 • Sun Ho Han1

Received: 5 December 2017 Ó Akade´miai Kiado´, Budapest, Hungary 2018

Abstract Isobaric interference is a major limitation of secondary ion mass spectrometry. We developed a simple correction method for polyatomic mass interferences from lead in isotope ratio measurements of uranium. Lead-generated isobars were measured to determine their formation rates relative to lead isotopes. The rates were used to mathematically subtract the isobar contributions from the signal intensities in the uranium mass range. The correction method was successfully verified using mixed uranium-lead samples (oxide powder and solution-dried residue). Keywords Secondary ion mass spectrometry  Uranium isotope analysis  Lead-induced isobaric interference

Introduction Secondary ion mass spectrometry (SIMS), in addition to inductively coupled plasma-mass spectrometry (ICP-MS) and thermal ionization mass spectrometry (TIMS), is a major workhorse for analyzing environmental samples for nuclear safeguards [1–4]. SIMS and TIMS have mainly been employed to analyze the isotopic compositions of individual nuclear particles. TIMS is combined with other techniques, such as secondary electron microscopy or fission track imaging preceded by neutron irradiation, because nuclear particles must be prescreened among the background particles [5, 6]. In contrast, SIMS solely allows both prescreening and isotopic ratio measurements to be performed in the same instrument, thanks to the ion imaging capabilities of the system combined with the Electronic supplementary material The online version of this article (https://doi.org/10.1007/s10967-018-5798-0) contains supplementary material, which is available to authorized users. & Jinkyu Park [email protected] 1

Nuclear Chemistry Research Division, Korea Atomic Energy Research Institute, 989-111 Daedeok-daero, Yuseong-gu, Daejeon 34057, Republic of Korea

2

Department of Radiochemistry and Nuclear Nonproliferation, University of Science and Technology, 217 Gajeong-ro, Yuseong-gu, Daejeon 34057, Republic of Korea

Automated Particle Measurement (APM) software. This prescreening program provides the location of uranium particles and an estimation of their isotopic composition before performing precise isotopic measurements of the individual particles [7]. Despite its outstanding performance, SIMS often suffers from mass interferences induced by background elements, such as Pb, W, and Ti, which are abundant in the environment [8]. These elements can combine with other elements to generate molecular isobars, e.g., 207-208Pb27Al, 182 54,56 W Fe, and 48-50Ti138 2 Ba. To discriminate uranium isotopes from these polyatomic interferences, the mass resolving power (m/Dm) must be larger than 3000. One solution is to employ large-geometry (LG) SIMS, which provides high secondary ion transmission at high mass resolving power, and multi-collection capabilities. However, LG-SIMS is not easily accessible because it is expensive and only a limited number of instruments are installed worldwide. Esaka et al. developed a pioneering method for minimizing isobaric interferences by coupling a micro-manipulation technique with TIMS [5]. They first searched for uranium-containing particles using scanning electron microscopy and then used a micro-manipulator to move the particles one-by-one for TIMS analysis (which is free from molecular isobaric interferences). This method provided an excellent performance; however, transferring the particles onto the TIMS filaments is a low-throughput process, and

123

Journal of Radioanalytical and Nuclear Chemistry

the particles could be lost during the transfer. An innovative version of this method utilized SIMS APM measurements to search for particles, but this method also suffered from the disadvantages of individually moving the particles for TIMS analysis [9]. We previously reported a ‘‘bypass’’ method to overcome interferences from gadolinium, molybdenum, and zirconium [10]. By measuring the signals from uranium dioxide ions, not uranium atomic ions, we could correct the errors that propagated from the interferences. This method, however, is only valid for major isotopes because molecular isobars greatly affect the mass ranges of dioxides of minor uranium isotopes. To tackle isobaric interferences in an alternative way, we adopted a simple subtraction method to remove the isobar contributions from the raw signal intensities in uranium mass ranges. The subtraction method is commonly used to correct the contribution from 235UH on the signal intensity of 236U [8]. The 235UH contribution is estimated from the 238UH formation rate, which is determined during the uranium isotope measurements. The 235U intensity is multiplied by the formation rate, and the resultant 235UH portion is subtracted from the 236 m/z signal to obtain the corrected 236U signal intensity. Here, we chose lead as the target interfering element. Lead is the 36th most abundant element in the earth’s crust and often coexists with nuclear elements because of its broad use in radiation-shielding materials [11, 12]. Lead can generate molecular isobars, such as PbAl, PbCO, and PbO2, that potentially interfere with uranium isotopes. We first measured the isobar formation rates for lead using SIMS. The isotopic ratios of prepared U/Pb mixed samples were analyzed; then, the lead contribution was subtracted from the signal intensities in the uranium mass range. Errors caused by the isobars could be corrected for major and minor isotopes. This method could be applied to specific samples containing lead and uranium.

Experimental PbO2 powders were deposited on 25-mm-diameter glassycarbon planchets (Hitachi Chemical Co., Ltd., Tokyo, Japan) using a vacuum-suction impact method [13]. To simulate uranium particles surrounded by a lead matrix, PbO2 and U3O8 (NBS CRM U020a) powders were deposited on an identical planchet at roughly a 1:3 ratio using the same method. Powder deposition was completed in glove bags, and the planchets were stored in clean staticdissipative plastic cases until the SIMS measurements. Solution-based samples were also prepared on a single carbon planchet. Aliquots of a standard Pb solution (5 lL, AccuStandard, 1000 ppm) and uranium nitrate solution

123

(5 lL, 1000 ppm, depleted uranium) were separately spotted and air-dried on the planchet to measure the isobar formation rates and uranium isotopic ratios. The Pb solution was mixed with the depleted uranium (DU) solution at U:Pb ratios of 1:10, 1:100, and 1:1000 (v/v); the mixed solutions (5 lL each) were also deposited at different positions on the same planchet following the above-mentioned procedure. SIMS was performed using a Cameca IMS 7f-Auto instrument, which is a small-geometry SIMS; the measurements were performed in microprobe and APM modes, and the specific measurement conditions are summarized in Table 1. Briefly, the Pb isotope signals were measured along with signals at 234, 235, 236, 238, and 239 m/z to determine the isobar formation rates. For the mixed samples, the signals were measured in an identical manner, and the correction was performed by subtracting the contributions of the isobars from the raw signals according to the following equations. Rx ðy Pbisobar Þ ¼ x I=I ðy PbÞ

ð1Þ

I ðx Ucorrected Þ ¼ I ðx Umeasured Þ I ðy Pbmeasured Þ  Rx ðy Pbisobar Þ ð2Þ In Eqs. (1) and (2), x = 234, 235, or 236 and y = 206, 207, or 208. I(xUcorrected) and I(xUmeasured) are the corrected and measured intensities, respectively, for the xU isotope, I(yPbmeasured) is the measured intensity for the yPb isotope, and Rx(yPbisobar) is the isobar formation rate corresponding to each Pb isotope at mass x (m/z). Rx(yPbisobar) is measured on the Pb-only sample. xI is the intensity at mass x, and I(yPb) is the intensity for each Pb isotope.

Results and discussion Isobar formation rates for a deposited PbO2 powder sample We prepared a PbO2 powder deposition sample to examine the levels of isobars generated by Pb. We obtained a mass spectrum and intensity profile with respect to time for Pb and U mass ranges for a chunk of PbO2 powder (Fig. 1). Intensity profiles were also collected at two different sample spots, and the results were used to calculate Pb isotopic ratios (Table 2) and isobar formation rates described later in this section. The mass spectrum demonstrated that the signals from the Pb isotopes are dominant, and significant signals are observed at 205 and 209 m/z, probably from Pb hydrides and background noise. The measured isotopic abundances of Pb vary slightly from the natural abundances of Pb isotopes (1–6% deviation) (Table 2), indicating that the hydride and background noise

Journal of Radioanalytical and Nuclear Chemistry Table 1 SIMS measurement conditions Automated particle measurement (APM) Primary ions

O2?

Primary acceleration voltage

? 15 kV

Secondary extraction voltage

? 5 kV

Impact energy

10 keV

Primary ion current

2 nA

Nominal mass resolution (10% peak height)

400

Raster size

500 lm

Mass sequence

208

Integration time

2s

Pb,

Microprobe measurement

20 * 50 lm

238

U

204

Pb, 206Pb, 207Pb, 208Pb, 234U, 235U, 236U, 238U, 238UH

2 s for Pb; 10 s for

substantially contribute to each isotopic intensity. In the mass range of U, signals with intensities of * 100 c/s are measured at 234, 235, 236, 238, and 239 m/z. These signals could interfere with major and minor uranium isotopes. The signal pattern at 234, 235, and 236 m/z resembles the pattern for the peaks at 206, 207, and 208 m/z, which correspond to the natural abundances of Pb isotopes. Therefore, we assume that the signals at 234, 235, and 236 m/z are mainly induced by 206Pb, 207Pb, and 208Pb, respectively. The actual intensity ratios of 234I/236I and 235 236 I/ I are slightly different from the ratios of 206 208 I/ I and 207I/208I (Table 2). Thus, the signals at 234, 235, and 236 m/z seem to each be derived from a single form of polyatomic specie, and other forms of molecules with minor contributions. Possible forms of isobaric species are 206-208Pb12C16O, 207-208Pb27Al, 206-208Pb28-30Si, etc., but the major contribution might originate from PbCO because there is a large supply of carbon and oxygen from the substrate and sputtering ion beam. Unlike that observed for other signals, those at 238 and 239 m/z drastically decrease as the measurement time increases, probably because of depletion of a component or components consisting of isobaric species during ion beam sputtering. The component(s) might originate from the surface-coated material that was used for the vacuumsuction impact method. The decay in the signals at 238 and 239 m/z could also be caused by preferential sputtering on the lighter elements relative to heavier elements [15]. Possible forms of isobaric interferences at masses 238 and 239 are 206-207Pb16O2, 208Pb30Si, 208Pb14N16O, and their hydrides. However, none of these isobars are significant enough to affect the isotopic analysis of uranium based on

234

U,

5 s for

235

2 s for

238

U,

236

U;

238

UH;

U

their decreasing trend and low intensities; therefore, we focused on the isobars generated at 234, 235, and 236 m/z. The level of isobar formation was determined by measuring the signals at 234, 235, and 236 m/z along with Pb isotope signals in SIMS microprobe mode. The isobar formation rates were calculated using Eq. (1) and averaged for 20 cycles per measurement. The calculated values for R234(206Pbisobar), R235(207Pbisobar), and R236(208Pbisobar) are 5.7(1) 9 10-4, 4.9(2) 9 10-4, and 4.4(1) 9 10-4, respectively. The measured formation rates are much higher than the formation rate of lead oxide or lead nitride (\ * 10-5) observed in a study using ICP-MS performed by Pollington et al. [16]. The authors commented that Pbinduced isobars negligibly affected uranium isotope analysis using ICP-MS. However, we observe a significant level of Pb-generated isobars using SIMS. This would negatively impact the accuracy of isotope ratio measurements for minor isotopes.

Correction of Pb interferences on uranium isotope measurements A powdered U–Pb mixed sample was used to examine the correction of Pb interferences in the uranium isotope data in a Pb-dominant environment. First, the sample was analyzed in a similar manner as the Pb-only sample (Fig. 2). Signatures of Pb and U isotopes are clearly seen in the mass spectra. The intensity profile of each isotope was then measured for 20 cycles at four different sample spots with a raster size of 20–50 lm, and the results were used to obtain the isotope data listed in Table 3. The sizes of the U and Pb particles varied from several microns to several tens of microns. We intentionally used large raster sizes to

123

Journal of Radioanalytical and Nuclear Chemistry

Fig. 1 PbO2 powder sample: SIMS mass spectra (top, Pb mass range; middle, U mass range) and intensity depth profiles for Pb isotopes and isobaric species arising in the mass range of U isotopes (bottom)

simultaneously monitor the mixed signals from both U and Pb particles. After the measurement, a cycle-by-cycle correction was attempted according to Eq. (2), using the isobar formation rates determined above. The resultant

Fig. 2 Powdered U–Pb mixed sample: SIMS mass spectra (top, Pb mass range; middle, U mass range) and intensity depth profiles for Pb and U isotopes (bottom)

intensities were used to obtain the uranium isotopic ratios, and the standard errors obtained from the corrected dataset of 20 cycles were used as uncertainties.

Table 2 Pb isotope data obtained for the PbO2 powder sample (n indicates the number of measurements) Pb isotopic abundance (%)

Intensity ratios 204

206

207

208

206

207

234 236

1.40 (1) 1.4

25.55 (3) 24.1

21.32 (1) 22.1

51.73 (3) 52.4

0.49

0.41

0.64

Pb

Measured (n = 3) Natural abundance [14]

123

Pb

Pb

Pb

Pb/08Pb

Pb/08Pb

I/

I

235 236

I/

0.45

I

Journal of Radioanalytical and Nuclear Chemistry Table 3 Uranium isotope data from a powdered U–Pb mixed sample (n indicates the number of measurements)

234

U/238U

235

U/238U

1.768 (3) 9 10-4

Certified

-4

236

U/238U

2.069 (1) 9 10-2 2.09 (3) 9 10

-4

1.204 (7) 9 10-4 1.83 (26) 9 10-4

As measured (n = 4)

2.20 (13) 9 10

Relative deviation from certified value

25%

1.1%

52%

Corrected

1.88 (6) 9 10-4

2.08 (4) 9 10-4

1.24 (1) 9 10-4

Relative deviation from certified value

6.3%

0.3%

2.9%

Fig. 3 Isotope ratio measurements for U–Pb mixed spots. Top, U:Pb = 1:10; middle, U:Pb = 1:100; bottom, U:Pb = 1:1000. Blue dashed lines indicate reference values confirmed by TIMS analysis. (Color figure online)

The as-measured data clearly demonstrated that the mass of the Pb-induced isobars interfered with the uranium isotopes, resulting in large deviations from the certified isotopic ratios, especially for minor isotopes (Table 3). After applying the interference correction, it efficiently reduced the relative deviation to 0.3% for 235U/238U and 3–6% for 234U/238U and 236U/238U; these are acceptable error ranges for uranium particle analysis using SIMS [8].

Application of the correction method to a sample with varying U:Pb ratios A solution-based approach was adopted for sample preparation to mimic lead-bearing uranium particles. An advantage of this approach is that we can easily manipulate the elemental composition of the sample, unlike the method used to prepare the powdered samples. To investigate the interference level at varying lead concentrations,

123

Journal of Radioanalytical and Nuclear Chemistry

Fig. 4 Boxplots of the isotope ratio measurements for U–Pb mixed spots. Top, U:Pb = 1:10; middle, U:Pb = 1:100; bottom, U:Pb = 1:1000. Red dashed lines indicate reference values confirmed by TIMS analysis. (Color figure online)

uranium and lead solutions were mixed at three different ratios that maintained the lead-dominant condition. The three mixed solutions were each spot-dried at different positions on the same planchet with U-only and Pb-only spots. DU was used as the uranium solution because a lower 235 U content facilitates monitoring the isobar contribution to the major isotopes. TIMS was used to confirm the isotopic ratios of this material because the DU solution is not a certified reference material. The isotopic ratios measured for the U-only spot by SIMS were compared with those obtained by TIMS; the results were consistent (Fig. S1). Isobar formation rates were re-examined on the Pb-only spot in the newly prepared solution-dried sample. Rates of 4.7(2) 9 10-4, 2.3(1) 9 10-4, and 2.0(1) 9 10-4 were obtained for R234(206Pbisobar), R235(207Pbisobar), and R236(208Pbisobar), respectively. These formation rates are on the same order of magnitude as those for the oxide powder.

123

This indicates that the pattern of isobar generation depends on the chemical composition and physical form. On the mixed sample spots, U and Pb were distributed unevenly on a micrometer scale (Fig. S2). Therefore, the ratios of U and Pb at local areas must also be highly varied. We randomly selected 15–20 different positions and obtained isotope profile data at a 50-lm raster size for each sample spot. The resultant data were corrected as described previously, and both the as-measured and corrected isotopic ratios were plotted for each measurement (Fig. 3). The presence of lead provided overestimated uranium isotopic ratios because of interferences. The degree of overestimation increased as the lead concentration increased. The 235U/238U signal is amplified by a factor of * 1.5 when a 1000-fold excess of lead is present in the sample. The signals for minor isotopes are augmented by one or two orders of magnitude. The isobaric interference also worsens the internal precision of each measurement.

Journal of Radioanalytical and Nuclear Chemistry

The large variations in the as-measured isotopic ratios indicate that uranium and lead are not evenly distributed at the microscale as observed in the secondary ion images. The interference correction was performed on the data following the same method as that applied for the solutionbased sample. The correction works well for 235U/238U under all three conditions and removes most of the interference contributions. However, the correction for minor isotopes is not excellent, except when U:Pb = 1:10. The scattered data points show large deviations and some negative values. Boxplots were used to comprehensively depict the scattered datasets (Fig. 4). For U:Pb = 1:1000, five data points with negative values were eliminated because negative isotopic ratios are not physically reasonable. Accordingly, the five data points with the highest values among this dataset were also excluded. The plots clearly exhibit that the correction was generally successful for 235 U/238U, reducing the ratios (i.e., near the reference value). 234U/238U and 236U/238U were ideally corrected for the spot with a tenfold excess of lead. The correction was not effective for the other two spots because of a high level of interference relative to the 234U and 236U signals; however, the corrected values are on the same orders of magnitude as the reference values.

Conclusions We developed a simple correction method for isobaric interferences induced by lead for uranium isotope analysis using SIMS. By determining the isobar formation rates, the subtractive correction successfully eliminated the isobaric contributions from signals in the uranium mass range. The efficiency of the correction method was clearly demonstrated using both powdered and solution-based samples. This method is applicable to samples that contain uranium with excess lead. Prescreening with other tools, such as X-ray fluorescence imaging which can quickly scan the elemental composition, could be used to indicate if the correction method is feasible. The approach is not applicable in case of specific lead isotopic composition in nuclear particles due to significant uranium decay. However, real-world samples are expected to contain lead originating mainly from shielding materials rather than from the uranium decay products. The content of lead in those samples would widely vary depending on where the samples are collected. If the samples are from a hot cell, lead could be present in a 1000-fold excess over uranium. This approach is not valid for a 1000-fold excess of lead, and not even for a 100-fold excess of lead when analyzing the minor isotopes. Further study is required to improve the

accuracy and precision of this method for applications to real-world samples. Acknowledgements We thank Ms. Ranhee Park for assistance with the TIMS measurements of the DU material used in this study. This work was supported by the Nuclear Safety Research Program through the Korea Foundation of Nuclear Safety (KOFONS), granted financial resource from the Nuclear Safety and Security Commission (NSSC), Republic of Korea (1405020).

References 1. Mayer K, Wallenius M, Ray I (2005) Nuclear forensics-a methodology providing clues on the origin of illicitly trafficked nuclear materials. Analyst 130(4):433–441 2. Tandon L, Hastings E, Banar J, Barnes J, Beddingfield D, Decker D, Dyke J, Farr D, FitzPatrick J, Gallimore D, Garner S, Gritzo R, Hahn T, Havrilla G, Johnson B, Kuhn K, LaMont S, Langner D, Lewis C, Majidi V, Martinez P, McCabe R, Mecklenburg S, Mercer D, Meyers S, Montoya V, Patterson B, Pereyra RA, Porterfield D, Poths J, Rademacher D, Ruggiero C, Schwartz D, Scott M, Spencer K, Steiner R, Villarreal R, Volz H, Walker L, Wong A, Worley C (2008) Nuclear, chemical, and physical characterization of nuclear materials. J Radioanal Nucl Chem 276(2):467–473 3. Mayer K, Wallenius M, Varga Z (2013) Nuclear forensic science: correlating measurable material parameters to the history of nuclear material. Chem Rev 113(2):884–900 4. Boulyga S, Konegger-Kappel S, Richter S, Sangely L (2015) Mass spectrometric analysis for nuclear safeguards. J Anal At Spectrom 30(7):1469–1489 5. Esaka F, Esaka KT, Lee CG, Magara M, Sakurai S, Usuda S, Watanabe K (2007) Particle isolation for analysis of uranium minor isotopes in individual particles by secondary ion mass spectrometry. Talanta 71(3):1011–1015 6. Konomi TE, Fumitaka E, Jun I, Kazunari I, Chi-Gyu L, Satoshi S, Kazuo W, Shigekazu U (2004) Application of fission track technique for the analysis of individual particles containing uranium in safeguard swipe samples. Jpn J Appl Phys 43(7A):L915 7. Peres P, Hedberg PML, Walton S, Montgomery N, Cliff JB, Rabemananjara F, Schuhmacher M (2013) Nuclear safeguards applications using LG-SIMS with automated screening capabilities. Surf Interface Anal 45(1):561–565 8. Ranebo Y, Hedberg PML, Whitehouse MJ, Ingeneri K, Littmann S (2009) Improved isotopic SIMS measurements of uranium particles for nuclear safeguard purposes. J Anal At Spectrom 24(3):277–287 9. Esaka F, Suzuki D, Yomogida T, Magara M (2016) Application of automated particle screening for effective analysis of individual uranium particles by thermal ionization mass spectrometry. Anal Methods 8(7):1543–1548 10. Park J, Kim TH, Lee C-G, Lee J, Lim SH, Han SH, Song K (2017) Combinatory use of time-of-flight secondary ion mass spectrometry (SIMS) and sector-field SIMS for estimating elemental and isotopic compositions of nuclear forensic samples. J Radioanal Nucl Chem 311(2):1535–1544 11. Lide DR (ed) (2005) Abundance of elements in the earth’s crust and in the sea. In: CRC handbook of chemistry and physics, Internet Version 2005 12. Schimpf W, MacKenzie DJ, Karson MZ, Schmidt BR, Casteras JE, Cullen JJ (2017) A guide to the use of lead for radiation shielding. Lead Industries Association, Inc. www.canadametal.

123

Journal of Radioanalytical and Nuclear Chemistry com/wp-content/uploads/2016/08/radiation-shielding.pdf. Accessed 29 Nov 2017 13. Esaka F, Watanabe K, Fukuyama H, Onodera T, Esaka KT, Magara M, Sakurai S, Usuda S (2004) 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(11):1027–1032 14. Meija J, Coplen TB, Berglund M, Brand WA, De Bie`vre P, Gro¨ning M, Holden NE, Irrgeher J, Loss RD, Walczyk T,

123

Prohaska T (2016) Isotopic compositions of the elements 2013 (IUPAC technical report). Pure Appl Chem 88(3):293–306 15. Smentkowski VS (2000) Trends in sputtering. Prog Surf Sci 64(1–2):1–58 16. Pollington AD, Kinman WS, Hanson SK, Steiner RE (2016) Polyatomic interferences on high precision uranium isotope ratio measurements by MC-ICP-MS: applications to environmental sampling for nuclear safeguards. J Radioanal Nucl Chem 307(3):2109–2115