Univariate and multivariate analyses of Gd in gadolinia-doped ceria using laser-induced breakdown spectroscopy

Abstract

We report laser-induced breakdown spectroscopy (LIBS) analysis of gadolinium (Gd) in gadolinium-doped ceria (GDC), the functional material used in solid oxide fuel cells (SOFCs). LIBS using laser ablation to atomize solid samples is competent in sampling the refractory ceria. Also, a few to 10 mol.% of Gd is doped in GDC to be used for SOFCs. Such a high concentration requires additional dilution after acid digestion of the sample for conventional analyses. LIBS with typical dynamic range from ppm to percent can analyze the high concentrations of Gd directly. GDCs with 1–10 mol.% of Gd were synthesized by glycine-nitrate process and the LIBS spectra were recorded. In spite of these advantages of LIBS, calibrating LIBS spectra to the concentrations of Gd is quite challenging due to rich emissions of Ce. Six prominent Gd I lines were selected for univariate calibrations and the whole spectra were used for multivariate partial least squares-regression (PLS-R) modeling. The PLS-R model considering emission intensities of both matrix and analyte elements outperformed the univariate calibrations. Our results suggest that LIBS in combination with multivariate analysis can be useful as an analysis method for the refractory lanthanide oxide materials.