Analysis of liquid sodium purity by laser-induced breakdown spectroscopy. Modeling and correction of signal fluctuation prior to quantitation of trace elements

Cécile Maury (1) , Jean-Baptiste Sirven (2) , Michel Tabarant (3) , Daniel L'Hermite (4) , J.-L. Courouau (5) , Catherine Gallou (6) , Nadège Caron (4) , Gilles Moutiers (6) , Valérie Cabuil (7)

1 LANIE - Laboratoire de développement Analytique Nucléaire Isotopique et Elémentaire
2 SCCME - Service de la Corrosion et du Comportement des Matériaux dans leur Environnement
3 CEA-DES (ex-DEN) - CEA-Direction des Energies (ex-Direction de l'Energie Nucléaire)
4 DPC - Département de Physico-Chimie
5 LECNA - Laboratoire d'Etude de la Corrosion Non Aqueuse
6 SEARS - Service d'études analytiques et de réactivité des surfaces
7 PECSA - Physicochimie des Electrolytes, Colloïdes et Sciences Analytiques

Liquid sodium is used as coolant in sodium-cooled fast nuclear reactors. Among many parameters to monitor to ensure the safe operation of the reactor, the coolant chemical purity is a relevant indicator of several undesirable situations, like corrosion of structural materials or sodium contamination, which may release different elements in the coolant. Several techniques have already been implemented to measure the sodium purity, but their response time is long and not suited for continuous monitoring. Therefore, as a complement to them, Laser-Induced Breakdown Spectroscopy (LIBS) is considered as a promising technique for real-time analysis of the coolant purity. In this paper we report on the first LIBS quantitative measurements performed in liquid sodium at 150℃. Calibration curves were traced for lead and indium using the standard addition method. Important intensity drifts and fluctuations were observed, mostly due to pressure variations in the sodium oven. Background subtraction and/or normalization was used to compensate for those intensity fluctuations. To describe the effect of these corrections on the analytical signal noise, a simple model was proposed and its results were found to satisfactorily fit the experimental data. Using this approach, the best detection limits were obtained for the background-subtracted and normalized data, and were found to be 6 ppm for lead and 5 ppm for indium.

Keywords: laser-induced breakdown spectroscopy; LIBS; liquid sodium; limit of detection; drift; noise; signal correction

Domains: Material chemistry; Chemical engineering

Cécile Maury, Jean-Baptiste Sirven, Michel Tabarant, Daniel L'Hermite, J.-L. Courouau, et al.. Analysis of liquid sodium purity by laser-induced breakdown spectroscopy. Modeling and correction of signal fluctuation prior to quantitation of trace elements. Spectrochimica Acta Part B: Atomic Spectroscopy, 2013, 82, pp.28-35. ⟨10.1016/j.sab.2012.12.005⟩. ⟨cea-02872388⟩