Home / Publications / Journals / Nuclear Technology / Volume 206 / Number 4
Nuclear Technology / Volume 206 / Number 4 / April 2020 / Pages 651-661
Technical Note / dx.doi.org/10.1080/00295450.2019.1670009
Articles are hosted by Taylor and Francis Online.
Cyclic voltammetry (CV) was used to study SmCl3 at concentrations of 0.42 to 8.99 wt% in molten eutectic LiCl-KCl (44.2:55.8 wt%) at 773 K. For each sample, CV was repeated at different electrode surface areas to measure the peak current density. By analyzing the measured peak current density and concentration relationship with the Randles-Sevcik equation, the Sm(III) diffusivity for each sample was calculated. These diffusion coefficients ranged from 0.934 × 10−5 to 1.572 × 10−5 cm2‧s−1, showing no noticeable trend with a change in concentration. The samples were then divided into two groups of five. The first group was used to develop a calibration model for concentration prediction, while the second group was used to test and validate the model. The first model was based on the relationship between current density and concentration. This model had a very low limit of detection of 0.14 wt% and very low error as evaluated by the root-mean-square error of calibration of 0.108 wt%. The second model was a multivariate approach utilizing the current density values and laser-induced breakdown spectroscopy (LIBS) intensities as regressors; however, the introduction of LIBS data showed an increase in the model’s prediction error when compared to the first model. The electrode withdrawal method proved to be a preferable option due to a substantial increase in precision.