Laser-induced breakdown spectroscopy; nuclear safeguards; pyroprocessing; molten salts;LIQUIDS

In the pyrochemical separation of used nuclear fuel (UNF), fission product, rare earth, and actinide chlorides accumulate in the molten salt electrolyte over time. Measuring this salt composition in near real-time is advantageous for operational efficiency, material accountability, and nuclear safeguards. Laser-induced breakdown spectroscopy (LIBS) has been proposed and demonstrated as a potential analytical approach for molten LiCl-KCl salts. However, all the studies conducted to date have used a static surface approach which can lead to issues with splashing, low repeatability, and poor sample homogeneity. In this initial study, a novel molten salt aerosol approach has been developed and explored to measure the composition of the salt via LIBS. The functionality of the system has been demonstrated as well as a basic optimization of the laser energy and nebulizer gas pressure used. Initial results have shown that this molten salt aerosol-LIBS system has a great potential as an analytical technique for measuring the molten salt electrolyte used in this UNF reprocessing technology.

Laser-induced breakdown spectroscopy; LIBS; safeguards; pyroprocessing; molten salt; cerium; gadolinium;MELT CRYSTALLIZATION; MOLTEN-SALT; STRONTIUM; CESIUM

Pyroprocessing of used nuclear fuel (UNF) has many advantages-including that it is proliferation resistant. However, as part of the process, special nuclear materials accumulate in the electrolyte salt and present material accountability and safeguards concerns. The main motivation of this work was to explore a laser-induced breakdown spectroscopy (LIBS) approach as an online monitoring technique to enhance the material accountability of special nuclear materials in pyroprocessing. In this work, a vacuum extraction method was used to draw the molten salt (CeCl3-GdCl3-LiCl-KCl) up into 4 mm diameter Pyrex tubes where it froze. The salt was then removed and the solid salt was measured using LIBS and inductively coupled plasma mass spectroscopy (ICP-MS). A total of 36 samples were made that varied the CeCl3 and GdCl3 (surrogates for uranium and plutonium, respectively) concentrations from 0.5 wt% to 5 wt%. From these samples, univariate calibration curves for Ce and Gd were generated using peak area and peak intensity methods. For Ce, the Ce 551.1 nm line using the peak area provided the best calibration curve with a limit of detection (LOD) of 0.099 wt% and a root mean squared error of cross-validation (RMSECV) of 0.197 wt%. For Gd, the best curve was generated using the peak intensities of the Gd 564.2 nm line resulting in a LOD of 0.027 wt% and a RMSECV of 0.295 wt%. The RMSECV for the univariate cases were determined using leave-one-out cross-validation. In addition to the univariate calibration curves, partial least squares (PLS) regression was done to develop a calibration model. The PLS models yielded similar results with RMSECV (determined using Venetian blind cross-validation with 17% left out per split) values of 0.30 wt% and 0.29 wt% for Ce and Gd, respectively. This work has shown that solid pyroprocessing salt can be qualitatively and quantitatively monitored using LIBS. This work has the potential of significantly enhancing the material monitoring and safeguards of special nuclear materials in pyroprocessing.