Fused glass sample preparation for quantitative laser-induced breakdown spectroscopy of geologic materials

Abstract

Laser-induced breakdown spectroscopy is a powerful analytical method, but LIBS is subject to a matrix effect which can limit its ability to produce quantitative results in complex materials such as geologic samples. Various methods of sample preparation, calibration, and data processing have been attempted to compensate for the matrix effect and improve LIBS precision. This study focuses on sample preparation by comparing fused glass as a preparation for powdered material to the more commonly used method of pressing powder into pellets for LIBS analysis of major elements in complex geologic materials. Pelletizing powdered material is a common and convenient method for preparing samples but problems with the physical matrix brought on by inconsistencies in the homogeneity, density, and laser absorption, coupled with the chemical matrix problem lead to spectral peak responses that are not always consistent with the absolute concentration of representative elements. Twenty-two mineral and rock samples were analyzed for eight major oxide elements. Samples were prepared under both glass and pellet methods and compared for internal precision and overall accuracy. Fused glass provided a more consistent physical matrix and yielded more reliable peak responses in the LIBS analysis than did the pressed pellet preparation. Statistical comparisons demonstrated that the glass samples expressed stronger separability between different mineral species based on the eight elements than for the pressed pellets and showed better spot-to-spot repeatability. Regression models showed substantially better correlations and predictive ability among the elements for the glass preparation than did those for the pressed pellets.