Development and validation of HPLC-SF-ICP-MS methodologies for characterization of spent nuclear fuel

Wanna Nancy

Promoter

Vanhaecke Frank, (UGent), Frank.vanhaecke@ugent.be

SCK•CEN Mentor

Van Hoecke Karen
karen.van.hoecke@sckcen.be
+32 14 33 82 01

SCK•CEN Co-mentor

Vasile Mirela
mirela.vasile@sckcen.be
+32 14 33 28 31

Expert group

Radiochemistry

PhD started

2017-01-15

Short project description

Over many years SCK-CEN has built up considerable expertise in spent nuclear fuel characterization and the determination of burn-up credit using experimentally obtained measurement data. This information is crucial, amongst others, for the validation of computational codes to calculate nuclide inventories, which is relevant to many aspects of nuclear plant operation [1].

Mass-spectrometry is one of the main analytical techniques that can provide information about the isotopic composition of (nuclear) samples. In order to obtain accurate data and the lowest limits of detection possible, spectral overlap of isobaric nuclides must be avoided by separating the individual elemental analytes. Until now, this has been achieved within the RCA group by using off-line gravimetric column chromatographic separations, which results in time-consuming sample preparation procedures. Once the elements have been isolated, the isotopic compositions and concentrations of the elements of interest are determined by means of thermal ionisation mass spectrometry (TIMS). Despite the high accuracy and precision obtainable with TIMS, its intrinsic characteristic, i.e. the need for off-line analysis of one analyte at a time, can be considered an important drawback.

The use of on-line high pressure liquid chromatography sector-field inductively coupled plasma mass-spectrometry (HPLC-SF-ICP-MS) could be an effective way to overcome the disadvantages of off-line isolation followed by TIMS analysis. In HPLC-SF-ICP-MS, the outlet of a chromatographic column is connected directly to the sample introduction system of a SF-ICP-MS instrument. The HPLC performs a rapid and automated separation of the analytes, while the SF-ICP-MS generates isotopic data. Although the best obtainable precisions with HPLC-SF-ICP-MS (2-4 % expanded uncertainty, k = 2) are inferior to those obtainable with TIMS (< 1 % expanded uncertainty, k = 2), this project can deliver a valuable scientific contribution in the evaluation of HPLC-SF-ICP-MS as a fit-for-purpose technique and in the comparison to TIMS. In a preliminary case study within the RCA group, the off-line quantitative separation of lanthanides by means of HPLC has already been demonstrated. However, through further optimization, the separation method will be applicable to both lanthanides and actinides in a variety of sample matrices. In the context of burn up determination, spent nuclear fuel samples are of primary interest. Secondly, environmental and biological samples will be investigated, aimed at achieving the lowest limits of detection possible and exploring the applicability of optimized methods for tracer studies in environmental and biological systems. Therefore, several aspects of HPLC-SF-ICP-MS methodology must be addressed. The HPLC operational parameters still need to be optimized for the relevant lanthanides and actinides, the HPLC eluent(s) and flow rate must be compatible with the SF-ICP-MS introduction system and the acquisition of accurate and precise isotope data from transient signals needs to be investigated.

The analytes of primary interest for spent fuel characterization and burn up calculation are U, Pu and Nd. Isotope dilution will be used as a calibration technique, as it results in smaller measurement uncertainties than external calibration or standard addition calibration, to make the technique as competitive as possible with TIMS. However, isotope dilution requires knowledge of the exact isotopic composition of a sample and a spike solution. For each analyte of interest, a measurement protocol needs to be developed in order to optimize the obtainable precision. For the analysis of environmental and biological samples, the main focus will be on U and Pu.

Objective

1. Optimization of the on-line separation of U, Pu and Nd from spent nuclear fuel by means of HPLC and validation using environmental and biological reference samples

To separate lanthanides and major actinides, a cation exchange column is used as the stationary phase. Multiple eluents containing inorganic acids and/or organic complexing agents are needed to sequentially elute Pu, U and Nd [2]. The separation method parameters need to be optimized in view of the column type (normal vs. microbore column) and eluent sequence, composition (isocratic vs. gradient elution) and flow rate in order to obtain interference-free analyte peaks that can be quantified by ICP-MS. In addition, the occurrence of matrix effects on analyte signals due to variable eluent composition must be investigated and corrected or prevented appropriately (e.g. use of additional gas, flow splitter, compensating gradient,…). The optimized on-line separation method(s) will also be applied to environmental and biological samples in order to investigate the lowest obtainable limits of detection possible for U and Pu. Hence, method validation and extending the method's applicability towards diverse sample matrices and analyte concentrations can be achieved.

2. Quantification of isotope ratios from transient signals and optimization of isotope dilution analysis

The chromatographic separation process results in Gaussian analyte peaks with a width of about 60 s. Within this time period, isotope ratios have been found to drift, resulting in the deterioration of isotope ratio precision [3]. Optimization of instrumental measurement parameters (i.e. the number of isotopes monitored within one run, dwell time, replicates, integration window,…) together with investigation of various isotope ratio calculation methods is intended to yield the best possible isotope ratio precisions. Examples of reported data reduction methods are peak area integration [4], point by point average isotope ratio [5] and linear regression slope method [6].

In addition, isotope ratio measurements must be corrected for mass discrimination which causes the measured isotope ratio to differ from the true isotope ratio, due to preferential extraction and transmission of heavier ions over lighter ones from the plasma. Correction using a standard of known isotopic composition, introduced in between sample injections (i.e. sample standard bracketing) [7, 8] or within one sample separation run via a separate inlet valve (intra injection sample standard bracketing) [9] is usually performed. Various mathematical correction models have been described and need to be evaluated using both these bracketing approaches. Analogously to the first objective, methods will be optimized for spent nuclear fuel sample matrices and will be validated using environmental and biological samples and reference materials to investigate method applicability for tracer studies.

3. Evaluation of measurement uncertainty and comparison with TIMS

As the uncertainty on the measurement results is an important criterion to evaluate their metrological quality, uncertainty calculations will be documented for different experimental and calculational approaches. Therefore, bottom-up error propagation rules as described in the GUM guidance document will be followed [10]. In this way, the results and their associated measurement uncertainties can be compared with the extensive data sets for U, Pu and Nd based upon TIMS analyses that are available within the RCA group.

References:

1.            Spent Nuclear Fuel Assay Data for Isotopic Validation. 2011, OECD.

2.            Günther-Leopold, I., et al., Measurement of plutonium isotope ratios in nuclear fuel samples by HPLC-MC-ICP-MS. International Journal of Mass Spectrometry, 2005. 242(2-3): p. 197-202.

3.            Rodriguez-Gonzalez, P., et al., Species-specific stable isotope analysis by the hyphenation of chromatographic techniques with MC-ICPMS. Mass Spectrom Rev, 2012. 31(4): p. 504-21.

4.            Krupp, E.M., et al., Precise isotope-ratio determination by CGC hyphenated to ICP-MCMS for speciation of trace amounts of gaseous sulfur, with SF6 as example compound. Analytical and Bioanalytical Chemistry, 2004. 378(2): p. 250-255.

5.            Epov, V.N., et al., Simultaneous determination of species-specific isotopic composition of Hg by gas chromatography coupled to multicollector ICPMS. Analytical Chemistry, 2008. 80(10): p. 3530-3538.

6.            Fietzke, J., et al., Alternative data reduction for precise and accurate isotope ratio determination via LA-MC-ICP-MS. Geochimica Et Cosmochimica Acta, 2008. 72(12): p. A267-A267.

7.            Gunther-Leopold, I., et al., Characterization of nuclear fuels by ICP mass-spectrometric techniques. Anal Bioanal Chem, 2008. 390(2): p. 503-10.

8.            Asai, S., et al., Isotope dilution inductively coupled plasma mass spectrometry for determination of Sn-126 content in spent nuclear fuel sample. Journal of Nuclear Science and Technology, 2013. 50(6): p. 556-562.

9.            Guéguen, F., et al., Neodymium isotope ratio measurements by LC-MC-ICPMS for nuclear applications: investigation of isotopic fractionation and mass bias correction. J. Anal. At. Spectrom., 2015. 30(2): p. 443-452.

10.          BIPM/IEC/IFCC/ISO/IUPAC/IUPAP/OIML, Evaluation of measurement data — Guide to the expression of uncertainty in measurement. 1995.

11.          Interne memo RCA/KvHo/15-10/. Actieplan : massaspectrometrische analyses RCA groep Nuclearisering Element 2 / Introductie van high pressure liquid chromatography (HPLC) als monstervoorbereiding / optimalisatie bepaling van isotopenverhoudingen van lanthaniden en actiniden met single- en multi-collector