Capabilities of high resolution ICP-OES for plutonium isotopic analysis Academic research paper on "Chemical sciences"

ELSEVIER

Contents lists available at ScienceDirect

Microchemical Journal

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

Capabilities of high resolution ICP-OES for plutonium isotopic analysis

Michael Krachler *, Rafael Alvarez-Sarandes

European Commission — Joint Research Centre, Institute for Transuranium Elements, P.O. Box 2340, D-76125 Karlsruhe, Germany

ARTICLE INFO ABSTRACT

The potential of a commercial high resolution (HR-)ICP-OES instrument for the reliable determination of the plutonium (Pu) isotopes 238Pu, 239Pu, 240Pu, 241Pu, and 242Pu in nuclear samples was investigated. Using a low flow, high efficiency, desolvating nebulizer enhanced the sensitivity of the employed HR-ICP-OES set-up several-fold, decreasing the Pu concentrations required for analysis to < 1 mg/kg, especially those of the highly radioactive 238Pu. To achieve both optimum optical resolution and highest sensitivity the width of the available apertures of the spectrometer was fine-tuned, yielding unsurpassed values at 20 ^im and 15 ^im for the entrance slit and exit slit, respectively. Appropriate Pu emission wavelengths for potential spectroscopic Pu isotopic studies were identified recording the emission spectra of mono-isotope solutions of 238Pu, 239Pu, and 242Pu, respectively. HR-ICP-OES spectra of 238Pu are reported for the first time, allowing not only the identification of the accurate position of the peak maxima of the individual Pu isotopes, but also the assessment of their peak width, peak shape and potential hyperfine splitting of emission signals. The largest isotope shift was observed for the Pu emission wavelength regions centered around \ = 402.154 nm, \ = 439.645 nm, and \ = 437.990 nm. The distance between 238Pu and 242Pu HR-ICP-OES signals expanded to 23 pm, 18 pm, and 17 pm for these Pu emission wavelengths. Well separated 238Pu or 239Pu and 242Pu emission signals were obtained at almost all investigated Pu emission wavelengths. In practice, however, peak deconvolution of the spectra is necessary because typical samples contain predominantly 239Pu and 240Pu whose emission signals overlap with each other, as demonstrated on selected samples originating from the institute's research projects related to the nuclear fuel cycle.

© 2015 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/).

CrossMark

Article history: Received 9 November 2015 Accepted 15 November 2015 Available online 29 November 2015

Keywords:

Plutonium

Isotopes

ICP-OES

High resolution

Nuclear

1. Introduction

Plutonium (Pu) inventories are steadily growing world-wide, e.g. some 70 tonnes of Pu contained in used fuel is removed when refuelling reactors each year. Numerous options are under investigation to make either use of the fissile isotopes in nuclear reactors or to develop proper conditioning and storage concepts. In all cases reliable analytical support is required to keep trustworthy control of the inventories. Pu is basically considered a man-made element comprising isotopes with long half-lives and high radiotoxicity. Among the recognized Pu isotopes, 238Pu (t1/2 = 87.74 y), 239Pu (t1/2 = 24,110 y), 240Pu (t1/2 = 6563 y), 241Pu (t1/2 = 14.4 y), and 242Pu (t1/2 = 373,000 y) are analyzed frequently, as the knowledge of their abundance in a specific sample provides strong indications on the origin and intended use of the element, i.e., weapons Pu, nuclear fuel, reprocessed Pu, and global fall-out [1-5]. The isotopes 238Pu, 239Pu, 240Pu, and 242Pu all decay by emission of a particles, while 241Pu undergoes (3 decay to produce 241Am (t1/2 = 432.2 y).

* Corresponding author at: European Commission — Joint Research Centre, Institute for Transuranium Elements, ITU, P.O. Box 2340, D-76125 Karlsruhe, Germany. Tel.:+ 49 7247 951 884; fax: +49 7247 951 588.

E-mail address: michael.krachler@eceuropa.eu (M. Krachler). URL: https://ec.europa.eu/jrc/en/institutes/itu (M. Krachler).

Mass spectrometric techniques such as [CP-MS, MC-1CP-MS and TIMS, but also a- and ^-spectrometry are commonly applied to determine these Pu isotopes in nuclear samples [2,3]. More recently, also accelerator mass spectrometry (AMS) is applied to Pu isotopic analysis of non-radioactive environmental samples [5,6]. The aforementioned analytical approaches normally require a chemical separation of the Pu from the other actinides present in the actual sample to reduce detrimental spectral interferences that would otherwise hamper the reliable determination of specific Pu isotopes. The quantification of 238Pu using ICP-MS, for example, suffers from the isobaric spectral interference of 238U (99.275% natural abundance), while the formation of 238UH deteriorates the reliable analysis of 239Pu using mass spectrometry [2,3]. Similarly, the presence of 241Am in a sample complicates the mass spectrometric measurement of the 241Pu. In a-spectrometry, in turn, the signals of 239Pu and 240Pu emerge as a single peak because both isotopes have similar emission energies which often cannot be discriminated with conventional instruments [3].

While all aforementioned, frequently employed and well established analytical techniques have some substantial drawbacks, optical emission spectrometry (OES) has rarely been used for Pu isotopic analysis. The first optical emission spectra of Pu date back more than 60 years [7]. The first analytical applications based on the use of OES employing an inductively coupled plasma as excitation source (ICP-OES), however,

http ://dx.doi.org/10.1016/j.microc.2015.11.028

0026-265X/© 2015 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/).

did not materialize before the 1980s [8,9]. These early studies already provided a huge amount of valuable information regarding the most prominent emission lines of Pu as well as their hyperfine interactions and isotopic line splitting which is based on electron-nuclear interactions [8,9]. It is the splitting of a specific Pu emission line into two or more signals that potentially provides useful Pu isotopic information. If such "isotope-specific" Pu emission lines can be separated from each other using ICP-OES, the analytical problems related to the mass and alpha spectrometric measurements mentioned above, could be readily overcome. It is worth noting that also attempts to determine 239Pu/240Pu ratios in nuclear samples using laser-induced breakdown spectroscopy (LIBS) operated at high resolution have been undertaken successfully [10].

In this vain, the current study attempts to elucidate the analytical potential of a commercial high resolution (HR)-ICP-OES instrument for the determination of 238Pu, 239Pu, 240Pu, 241Pu and 242Pu in samples related to the nuclear fuel cycle. Specifically, emission spectra of individual Pu isotopes are investigated to study the shape of obtained emission signals and the magnitude of the isotope shift at various Pu emission wavelengths. The analysis of Pu isotopes in actual sample solutions stemming from the nuclear fuel cycle highlights the potential and limitations of the proposed analytical procedure.

2. Experimental

2.1. instrumentation

All measurements were carried out with a commercial high resolution (HR-)ICP-OES instrument (Ultima2, HORIBAJobin Yvon, Longjumeau, France). The entire sample introduction system including the autosampler (AS500, HORIBA) of this sequentially working optical spectrometer was installed in a glove box enabling the analysis of radioactive samples. This demanding adaptation of the original ICP-OES setup protected the analyst from a and (3 radiation of the investigated radioactive sample solutions. A PolyPro ST nebulizer (Elemental Scientific, Inc., Omaha, NE, USA), pumped at a sample uptake rate of ~0.3 ml/min, was attached to a high efficiency sample introduction system (Apex Q, Elemental Scientific, Inc.,) thereby enhancing ICP-OES performance. A 40.68 MHz radio frequency generator operated at 1000 W forward power provided a robust plasma for sample analysis. All emission spectra presented here were recorded applying an integration time of 1 s/ measurement point.

2.2. Reagents, reference materials and samples

High purity water (18.2 Mfi cm) from a MilliQ Element (Millipore) water purification system and sub-boiled nitric acid were used for the preparation of all solutions.

Because of their radioactive nature, all investigated solutions were handled in appropriate glove boxes following strict safety and security procedures. As such, all analyzed samples had to be bagged-in into a dedicated glove-box through a lock system prior to any measurement. To minimize the radioactive dose to the operator, liquid samples were fed to the sample introduction system of the ICP-OES with an autosampler.

A concentrated Pu solution was prepared by dissolving a piece of Pu metal (~0.55 g; Pu content: 99.90 ± 0.04%) in hydrochloric acid as recommended by the supplier of this reference material (Plutonium Metal "MP2", CETAMA, CEA VALHRO, Marcoule, France). Isotopic information values indicated 97.76 mass% of 239Pu and 2.20 mass% of 240Pu accounting for more than 99.9% of the Pu in this specific reference material. This standard served as 239Pu standard solution.

In addition, well-characterized mono-isotopic solutions of both 238Pu [11] and 242Pu were employed and have been tested thoroughly in-house for potential impurities prior to analysis. Mono-isotopic solutions of 240Pu and 241Pu were not available for this study.

The following Pu solutions related to the nuclear fuel cycle were investigated: (1) standard reference material SRM 949f Plutonium Metal prepared by Los Alamos National Laboratory; and (2) several reference solutions whose Pu isotopic composition has been assessed in-house using TIMS. These solutions covered a wide variety of different Pu isotopic compositions, i.e., mainly 239Pu, mixtures of 239Pu and 240Pu as well

as samples containing 238Pu, 239Pu, 240Pu, 241Pu, and 242Pu.

3. Results and discussion

In contrast to the limited previous work on quantitative Pu analysis (Pu concentration and its individual isotopes) using ICP-OES [8,9], we aimed at lowering the sample amount necessary for Pu analysis by at least one order of magnitude helping to reduce both the radiological dose to the operator of the analytical instrument and the amount of radioactive waste generated throughout the analysis sequence.

3.1. Optimum instrumental resolution

The distance between two Pu emission wavelengths of various Pu isotopes, i.e., the isotopic splitting or frequently also referred to as isotope shift of Pu, spreads only a few picometers [7-9]. Conventional ICP-OES instruments typically provide an optical resolution in the range of 10-20 pm, while this value is well below 5 pm for most Pu emission wavelengths using HR-ICP-OES [12]. Therefore, the use of a high resolution spectrometer operated at its optimum optical resolution setting is an unequivocal prerequisite for this kind of study.

Initially, the available settings of both the entrance slit and the exit slit of the employed optical spectrometer were varied to identify the best compromise between optimum sensitivity and optimum resolution. To this end, both settings of each the entrance slit (10 pmor 20 |m) and the exit slit (15 |m or 80 |m) and combinations thereof were tested aspirating a solution containing ~3.5 mg/kg Pu, with Pu being predominantly present as 239Pu (~65%) and 240Pu (~26%). The outcome of this experiment is summarized numerically in Table 1 as well as displayed graphically in Fig. 1 for four representative Pu emission wavelengths.

Although opening the aperture of the entrance slit from 10 |m to 20 |m had only a minor impact on the achievable optical resolution, net peak intensities roughly doubled using the larger slit width

Table 1

Influence of slit width on the spectroscopic characteristics of HR-ICP-OES signals at four representative Pu emission wavelengths.

Slitsa Wavelength, nm

381.488 393.555 402.154b 453.615

Peak width, pmc

1015 6.96 5.53 4.55 8.14

2015 6.79 5.53 4.27 6.45

10 80 7.04 10.9 12.3 28.2

20 80 11.9 10.2 12.5 28.2

Net peak intensity, counts per fig/L

1015 151 503 165 334

20 15 291 1013 334 569

10 80 616 1892 547 631

20 80 1183 3674 1115 1213

Signal-to-background ratio

1015 3.11 65.6 3.90 15.1

20 15 4.11 92.3 5.80 15.7

10 80 3.91 58.1 4.50 6.68

20 80 4.32 59.3 5.08 7.10

The employed solution contained ~2.5 mg/kg 239Pu and ~1 mg/kg 240Pu, i.e., a total of ~3.5 mg/kg Pu.

a Slit widths [|jm] of the spectrometer: the first value refers to the aperture of the entrance slit, while the second value indicates the width of the exit slit. b Only values for the first peak i.e., 239Pu, are reported in the table, see text for details. c Peak width expressed as full width at half maximum (FWHM).

wavelength, nm wavelength, nm

Fig. 1. Influence of entrance (10 |jm or 20 |jm) and exit slit (15 |jm or 80 |jm) aperture of the employed HR-ICP-OES on the instrumental response at selected Pu emission lines. Spectra highlight the change of peak shape, optical resolution and intensity of a Pu solution containing ~2.5 mg/kg 239Pu and ~1 mg/kg 240Pu at (A) K = 381.488 nm, (B) K = 393.555 nm, (C) K = 402.154 nm, and (D) K = 453.615 nm.

(Table 1). In addition, the signal-to-background ratio improved when employing the 20 |m entrance slit setting. Interestingly, the position of the top of the Pu peaks was obviously also shifted when varying the width of the entrance slit (Fig. 1).

Switching the exit slit of the spectrometer from 15 |m to 80 |m resulted in a substantial loss of optical resolution yielding broad Pu emission signals (Fig. 1). For example, the peak width — defined as full width at half maximum (FWHM) — of the Pu signal at \ = 393.555 nm doubled when opening the exit slit to 80 |m. At \ = 402.154 nm and \ = 453.615 nm, the situation was even worse yielding 3- to 4-fold broader peaks compared with the narrower slit setting (Table 1). The only positive consequence of using a broader exit slit aperture was a gain in peak intensity (Fig. 1). The signal-to-background ratio, however, generally did not improve, but quite the opposite, it deteriorated substantially (Table 1).

In summary, the best compromise between optimum optical resolution and high sensitivity, both of which are essential for potential Pu iso-topic analysis, are the following instrumental aperture settings: entrance slit: 20 |m and exit slit: 15 |m.

3.2. Selection ofPu wavelengths

In addition to the technical prerequisites that provide the required optical resolution of emission signals for the unequivocal identification of individual Pu isotopes, it is frequently beneficial to select those Pu emission wavelengths that provide both largest isotopic splitting and narrowest peaks. In a recent study [13] we have characterized suitable Pu emission wavelengths for reliable Pu concentration analysis of nuclear samples using HR-ICP-OES. Method optimization focused on identifying Pu emission wavelengths that are sensitive and selective, i.e., free from spectral interferences of potential concomitant elements in the analyte solution such as Am, Np, Th and U [13].

Experience has shown that emission wavelengths that are best suited for concentration analysis, frequently do not work adequately for iso-topic work as already demonstrated for U isotopic analysis using HR-ICP-OES [14,15]. While for concentration analysis emission wavelengths with no or only marginal isotope shift are advantageous, isotopic analysis requires an isotope shift of the target isotopes as large as possible. It is only then that emission signals of various isotopes of the same element can be optically resolved, identified, and quantified properly.

Similar to the spectrochemical behavior of U [14,15], the statement made above also applies to Pu. Two (\ = 449.378 nm and \ = 453.615 nm) of the eleven Pu emission wavelengths identified for potential isotopic Pu analysis in this study (Table 2) were already selected earlier for the reliable quantification of total Pu, i.e., the Pu concentration [13]. These two Pu emission wavelengths have the smallest isotope shift among the tested wavelengths (Table 2). In other words, even

Table 2

Isotope shift of 238Pu, 239Pu, and 242Pu as identified using HR-ICP-OES.

Shift [pm] Wavelength [nm] Shift [pm] Total shift [pm]

238Pu-239Pu 239Pu 239Pu—242Pu 23sPu-242Pu

2 381.488 — 7 9

-3 390.721 9 12

-3 397.220 12 15

398.988 _* _*

-6 402.154 17 23

-1 435.271 6 7

4 437.990 -13 17

-4 439.645 14 18

-3 447.270 6 9

-1 449.378 6 7

-2 453.615 3 5

* Because of the hyperfine splitting of individual Pu emission lines (Fig. 3) and necessary peak deconvolution no information is reported.

though various Pu isotopes might be present in an actual sample, the small isotope shift will only marginally influence the wavelength position of the peak maximum. Therefore these two Pu emission lines are well suited for the accurate assessment of the Pu concentration of a sample, but reveal only a low potential for Pu isotopic work.

In contrast, Pu concentration analysis at K = 439.645 nm frequently led to diminished results and therefore was excluded for quantification purpose in our previous study [13]. The current investigation details the reason for this earlier reported (mis-)behavior of the Pu emission wavelength at K = 439.645 nm (Table 2). The isotope shift at this wavelength is ~3-times larger than that observed at K = 449.378 nm and K = 453.615 nm (Table 2). As a consequence, the intensity of the nominal 239Pu ICP-OES signal will be lowered if an investigated sample contains other Pu isotopes than 239Pu, which will produce additional peaks close to the prominent 239Pu emission wavelength. In fact, samples analyzed in our earlier study [13] contained up to ~10% 240Pu, which easily explains the lower concentrations calculated at K = 439.645 nm. Therefore the use of K = 439.645 nm should be avoided for Pu concentration analysis, but this wavelength domain provides excellent potential for the determination of Pu isotopes using HR-ICP-OES.

The superior optical resolution capabilities of the employed HR-ICP-OES instrument delivers the technical potential for identifying suitable wavelengths for monitoring the isotopic line splitting of Pu. Below we discuss in detail the choice of and criteria for selecting various Pu emission wavelengths for potential isotopic analysis.

3.3. Assessment of the isotopic splitting ofPu

Earlier work on high resolution optical emission spectra of Pu already identified worthwhile wavelength regions for Pu isotopic analysis

emission wavelength, nm

emission wavelength, nm

[7-9]. As such, the study of Edelson and co-workers [9] served as basis for further in-depth investigations focusing on those Pu emission wavelengths having the largest isotope shift. To reassure previously reported data, emission spectra were recorded using single isotope solutions of both 239Pu and 242Pu at a concentration of ~1 mg/kg each, together with a blank solution containing 1% HNO3. Some potentially useful Pu emission lines (e.g., K = 412.350 nm) reported earlier [9] were nottest-ed thoroughly in the current study because their sensitivity was too low to be of any practical relevance for the analysis of our actual samples.

For some other emission wavelengths our experimental findings disagreed partly with previous conclusions (Fig. 2). While the earlier described isotopic splitting of~11 pm [9] between 239Pu and 242Pu could be confirmed at K = 385.685 nm in the current study, a huge non-Pu-specific signal overlapped with the 242Pu emission signal (Fig. 2A). Both the blank solution and the 239Pu standard solution revealed a substantial ICP-OES signal located exactly at the spectral position of the 242Pu emission wavelength. If this non-specific spectral background was subtracted from the Pu emission spectra, the Pu isotope shift would become more obvious.

Such an exercise was performed at K = 391.348 nm where a similar non-Pu-specific spectral background was observed (Fig. 2B). As a result, two baseline separated peaks, namely those of 239Pu and 242Pu, became evident in the background corrected spectra (Fig. 2B). However, the high uncertainty associated with the huge blank subtraction that was needed for both Pu emission wavelengths at K = 385.685 nm and K = 391.348 nm did not favor them for accurate isotopic Pu analysis. The complex splitting of both 239Pu and 242Pu into at least three peaks each and more importantly, their overlapping nature also disqualified K = 398.998 nm for further in-depth investigations (Fig. 2C). Interestingly and contradicting earlier studies [9], no Pu-specific ICP-OES signal

emission wavelength, nm

emission wavelength, nm

Fig. 2. HR-ICP-OES spectra highlighting severe spectral interferences and hyperfine splitting thereby excluding the use of four potential emission wavelengths for Pu isotopic analysis. Displayed spectra represent single isotope solutions of both 239Pu and 242Pu at a concentration of ~1 mg/kg each as well as a blank solution containing 0.14 M HNO3. (A) K = 385.685 nm; (B) K = 391.348 nm; (C) K = 398.988 nm; (D) K = 419.006 nm. See text for details.

could be established in the vicinity of the proposed Pu emission line centered around \ = 419.006 nm (Fig. 2D). Taken together, all four wavelength regions discussed above were excluded from further examination because they either suffered from severe non-Pu-specific emission or revealed no Pu-related ICP-OES signal at all.

Despite this moderate disagreement with previous data, eleven Pu emission wavelengths proved useful for further in-depth investigation. To this end, emission spectra of 1 mg/kg mono-isotope solutions of each 238Pu, 239Pu, and 242Pu were recorded at the eleven wavelengths listed in Table 2. The values in the second column of this table refer to the well-established reference emission wavelength of the 239Pu isotope. Numbers in columns 1 and 3 specify the extent and direction of the isotope shift of the emission lines of 238Pu and 242Pu, respectively, relative to 239Pu. Negative numbers indicate a shift towards lower wavelengths and vice versa. While emission lines are normally moved upwards with increasing mass number of the Pu isotopes (Table 2), there are two exceptions, namely \ = 381.488 nm and \ = 437.990 nm. At these two wavelength regions the emission line pattern of the investigated Pu isotopes is inverted (Fig. 3).

The largest Pu isotope shift was found at \ = 402.154 nm, followed by \ = 439.645 nm and \ = 437.990 nm (Table 2). At \ = 402.154 nm the distance between 238Pu and 242Pu amounted to 23 pm, providing the potential for a sound separation of the emission lines of the individual Pu isotopes. However, even though the spectral background was quite clean, small satellite peaks were evident about 15-20 pm right of the actual major emission line of each Pu isotope (Fig. 3). These satellite peaks may deteriorate the determination of 241Pu and 242Pu in samples containing these two Pu isotopes using this specific emission wavelength region.

Among the other investigated wavelength regions producing a sufficiently large isotope splitting, those centered around \ = 397.220 nm and \ = 439.645 nm provided best potential for Pu isotopic analysis (Fig. 3). Either the insignificance of the actual isotope shift and/or a noisy spectral background excluded the remaining wavelength regions for further detailed investigations (Fig. 3).

3.4. Comparison with previous studies

Even though there exists some minor disagreement between earlier work [9] and the current study (see Section 3.3), the investigation of Edelson and co-workers served as an excellent starting point for further in-depth explorations of Pu isotopic analysis using HR-ICP-OES. However, because no blank spectra were reported in [9], non-Pu-specific emission signals could not be assessed. This lack of information might have led to wrong interpretations of the spectra previously as detailed in Fig. 2 and therefore stresses the importance of additionally recording blank spectra.

While a hyperfine splitting of the 239Pu emission signal at \ = 453.615 nm was observed by Edelson et al. [9], this splitting could not be reproduced at all in the current study (Fig. 3). The reason for this discrepancy remains unclear at present, however, it is worth noting that the disagreement cannot be related to the different monochromators used in both studies. The 1.5 m focal length of the monochromator used in the earlier investigation [9] provided a practical optical resolution (FWHM) of 2.5 ± 0.3 pm at \ = 402.154 nm. Employing a mono-chromator with 1 m focal length in the current survey deteriorated this value to ~4.3 pm. Even though there is an almost twofold difference in optical resolution between the two studies, this feature is too small to explain the above contrarious findings.

Due to the application of a high efficiency sample introduction system and an advanced photomultiplier providing improved sensitivity, the Pu concentration of individual isotopes required for analysis was lowered more than 200-times, i.e., from ~200 mg/kg [9] to < 1 mg/kg in the current study. This experimental approach helps reduce both the radioactive dose to the analyst and the amount of radioactive waste that is increasingly expensive to dispose of.

Generally speaking, the isotope shifts reported earlier [7-9] for 239Pu-242Pu agreed reasonably well with the values obtained in the present study (Table 2), even though there is still room for improvement in terms of both accuracy and precision of emission wavelength positions and the corresponding isotope shift between the various Pu isotopes. For the first time ever, optical emission spectra of 238Pu emitted from an ICP are presented in the current investigation. The inclusion of 238Pu in our investigation not only yielded new data on accurate 238Pu emission wavelengths (Table 2), but also provided new insights into peak shapes of this specific Pu isotope (Fig. 3).

3.5. Analysis ofnuclear samples

Among the investigated Pu emission lines, the ICP-OES response centered around \ = 402.154 nm provided both high sensitivity and the largest isotopic line splitting of the Pu emission signals. As such, all experiments described below, were conducted in the vicinity of the nominal 239Pu emission wavelength \ = 402.154 nm. At this Pu emission wavelength region the optical resolution of the employed HR-ICP-OES is 4.3 pm which is the technical prerequisite to resolve the Pu emission lines of individual Pu isotopes. Consequently, to test the potential of HR-ICP-OES for Pu isotopic analysis, several real-world samples were investigated at \ = 402.154 nm. Two examples of these experiments are summarized in Fig. 4, together with emission spectra of a solution containing only ~0.3 mg/kg of each 238Pu, 239Pu, and 242Pu. These reference spectra of individual Pu standard solutions are intended to support the identification of the Pu isotopes present in the "unknown" samples.

3.5.1. Standard reference material SRM 949f Plutonium Metal

The spectrum of a dissolved aliquot of the standard reference material SRM 949f Plutonium Metal is displayed in Fig. 4A. This solution comprises ~97.1 at.% 239Pu and ~2.8 at.% 240Pu, the sum of all other Pu isotopes remaining well <0.1 at.% in this reference material. The emission spectrum of this sample highlights the dominant 239Pu signal, as expected. However, another small peak is also clearly noticeable at \ = 402.168 nm (Fig. 4A). Even though a comparison of this ICP-OES signal to the reference spectrum of a mono-isotopic 242Pu standard solution might indicate that this emission peak refers to 241Pu or 242Pu, it does not, mainly because of two reasons: first, SRM 949f contains only minute amounts (<0.1 at.%) of 241Pu and 242Pu that cannot be detected with the currently applied analytical approach. Second, to exclude any potential contamination with another Pu isotope, TIMS analysis of this sample was conducted in-house, confirming the Pu isotopic composition reported on the certificate of SRM 949f. It is also important to note that the emission signal seen at \ = 402.168 nm also cannot be attributed to the 240Pu present in the standard reference material. The current peak is both at the wrong wavelength position (should be centered around \ = 402.161 nm if it was 240Pu) and additionally the ICP-OES response is far too substantial to account for only ~2.8 at.% 240Pu. The actual 240Pu signal at \ = 402.161 nm only becomes visible, if one zooms into the current spectrum (not shown). This approach would then also indicate that the 240Pu emission peak is suffering from spectral overlap of the huge 239Pu signal, rendering a quantification of the 240Pu difficult under the current experimental conditions.

In fact, the noticeable emission signal at \ = 402.168 nm represents a satellite peak of the main 239Pu signal (see also Fig. 3). Altogether, the abovementioned facts stress the importance of assessing both the emission spectra and the accurate wavelength position of each individual Pu isotope carefully to avoid any misinterpretation of the experimental data.

3.5.2. Spent nuclear fuel

The second example presented here comprises the Pu fraction of a high burn-up spent nuclear fuel containing ~1 wt.% 238Pu, ~65 wt.%

381.48 381.49

wavelength, nm

.¡? 60000 -

390.72 390.73

wavelength, nm

397.22 397.23

wavelength, nm

398.97 398.

398.99 399.00 399.01 399.02

wavelength, nm

70000 -

402.14 402.16 402.18

wavelength, nm

435.26 435.2

wavelength, nm

O 120000-

437.98 438.00 438.02

wavelength, nm

439.60 439.62 439.64 439.66

wavelength, nm

447.26 447.2

wavelength, nm

449.38 wavelength, nm

453.58 453.60 453.62

wavelength, nm

Fig. 3. High resolution emission spectra of single isotope solutions containing ~1 mg/kg of each Pu, Pu and Pu together with a 0.14 M HNO3 blank solution recorded at eleven wavelengths providing extensive Pu isotope splitting.

150000 -

160000

160000

120000 -

120000

120000

90000 -

30000 -

150000

120000

120000

105000 -

160000

140000-

125000 -

120000

100000 -

100000-

75000 -

80000-

50000 -

140000-

160000 -

120000 -

105000-

80000 -

70000-

40000 -

80000 -

2 60000

œ 40000 -

'"> 20000

Pu-239

Pu-238

Pu-242

402.14 402.15 402.16 402.17 wavelength, nm

402.18

tn 60000

'w 40000 en

Pu-239

Pu-238

Pu-242

402.14

402.15

402.16 402.17 wavelength, nm

402.18

0 20000

te 10000

c 8000

n TO 6000

4% Pu-241

4% Pu-242

402.14 402.15 402.16 402.17 wavelength, nm

402.18

Fig. 4. High resolution emission spectra of selected samples related to the nuclear fuel cycle. Colored spectra of solutions of -0.3 mg/kg of each 238Pu, 239Pu, and 242Pu are displayed for a comparison with and improved identification of Pu emission signals in the actual samples. HR-ICP-OES spectra of (A) the standard reference material 949f Plutonium Metal containing 239Pu and 240Pu; (B) a high burn-up spent fuel solution comprising 238Pu, 239Pu, 240Pu,241 Pu, and 242Pu, and (C) spectrum of (B), but displaying signal intensities on the y-axis on a logarithmic scale to highlight the ICP-OES response due to the presence of individual Pu isotopes in this specific sample. See text for further details.

239Pu, ~26 wt.% 240Pu, and ~4 wt.% of each 241Pu and 242Pu (Fig. 4B). Again, emission spectra of 238Pu, 239Pu and 242Pu standard solutions are displayed together with the emission data of the spent fuel for a better identification of the individual Pu isotopes. The dotted line in Fig. 4B

represents an additional spectrum of the actual spent fuel solution recorded with a narrower aperture of the spectrometer (entrance slit: 10 |jm, exit slit: 15 |jm) compared with the optimum setting of 20 |am and 15 |jm identified in Section 3.1. This spectrum proves that the use of a smaller entrance slit improves the achievable resolution of the employed spectrometer only marginally, for the sake of reduced signal intensities.

Both 239Pu and 240Pu are clearly evident in the emission spectrum of this spent fuel solution (Fig. 4B). However, the minor amounts of 238Pu, 241Pu, and 242Pu are not unequivocally identifiable in the spectrum. While the small 238Pu signal is buried under the intensive 239Pu signal, 241Pu and 242Pu emerge together as a broad shoulder right to the recognizable emission signal of 240Pu (Fig. 4B). As such, isotopic information on the minor abundant Pu isotopes is not extractable easily from the current spectra. In order to highlight the before mentioned difficulties, Fig. 4B was plotted another time with ICP-OES signal intensities now on a logarithmic scale (Fig. 4C). This supplementary figure indicates clearly the presence of both 241Pu and 242Pu in the spent fuel represented by a broad shoulder having two maxima. Furthermore, an additional peak centered around \ = 402.178 nm becomes immediately evident, a fact hardly seen previously in Fig. 4B. This specific ICP-OES signal suggests the presence of an interfering analyte other than a Pu isotope in the investigated spent fuel (Fig. 4C).

4. Conclusions

Individual emission spectra of mono-isotopic 238Pu, 239Pu, and 242Pu standard solutions demonstrate the potential of HR-ICP-OES for Pu isotopic analysis. Due to the small isotope shifts between neighboring Pu isotopes, however, the achievable optical resolution of 4.3 pm limits its application to genuine samples. Only careful assessment of detailed emission spectra of mono-isotopic standard solutions together with appropriate peak deconvolution strategies will pave the way for quantitative Pu isotopic analysis using HR-ICP-OES. Additionally, commercial ICP-OES instruments providing improved optical resolution will ease the interpretation of Pu emission spectra. For the time being, current experimental conditions are limited to a qualitative elucidation of the Pu isotopes present in a nuclear sample.

Acknowledgments

The supply of selected sample solutions, the employed certified reference material as well as the 238Pu, 239Pu and 242Pu mono-isotopic Pu standard solutions by our colleagues C. Apostolidis, A. Nicholl, M. Vargas Zuniga, and G. Rasmussen at EC-JRC-ITU is greatly appreciated.

References

[1 ] M.E. Ketterer, K.M. Hafer, J.W. Mietelski, J. Environ. Radioact. 73 (2004) 183-201.

[2] M.E. Ketterer, S.C. Szechenyi, Spectrochim Acta B 63 (2008) 719-737.

[3] X. Hou, P. Roos, Anal. Chim. Acta 608 (2008) 105-139.

[4] F. Luisier, J.A Corcho Alvarado, P. Steinmann, M. Krachler, P. Froidevaux,J. Radioanal. Nucl. Chem. 281 (2009) 425-432.

[5] P. Steier, E. Hrnecek A. Priller, F. Quinto, M. Srncik, A. Wallner, G. Wallner, S. Winkler, Nucl. Instrum. Methods Phys. Res. Sect. B 294 (2013) 160-164.

[6] F. Quinto, E. Hrnecek M. Krachler, W. Shotyk, P. Steier, S.R Winkler, Environ. Sci. Processes Impacts 15 (2013) 839-847.

[7] J.G. Conway, M. Fred, J. Opt. Soc. Am. 43 (1953) 216.

[8] E. Mainka, H.G. Müller, F. Geyer, Report KfK 3578 (in German), "Versuche zur Pu-Bestimmung mit Hilfe der ICP-Methode", Kernforschungszentrum Karlsruhe. , Institut für Radiochemie, Karlsruhe, Germany, 1983.

[9] M.C. Edelson, E.L. DeKalb, R.K. Winge, V.A Fassel, Spectrochim. Acta B 41 (1986) 475-486.

[10] C.A Smith, MA. Martinez, D.K. Veirs, D.A Cremers, Spectrochim. Acta B 57 (2002) 929-937.

[11] B. Kanellakopulos, J.I. Kim, E. Dornberger, Fresenius' Z. Anal. Chem. 323 (1986) 818-820.

[12] M. Krachler, R Alvarez-Sarandes, S. Van Winckel, J. Radioanal. Nucl. Chem. 304 (2015) 1201-1209.

[13] M. Krachler, R Alvarez-Sarandes, J. Anal. At. Spectrom. 30 (2015) 1655-1662.

[14] M. Krachler, P. Carbol, J. Anal. At. Spectrom. 26 (2011) 293-299.

[15] M. Krachler, D.H. Wegen, J. Anal. At. Spectrom. 27 (2012) 335-339.