A statistical definition of limit of detection for calibration-free laser-induced breakdown spectroscopy

Abstract

Laser-Induced Breakdown Spectroscopy (LIBS) is a fast and powerful technique for elemental analysis, but its high limit of detection (LOD) can limit its use, particularly in the analyses of trace elements. Although the LOD has been established for calibration curves and multivariate models in LIBS, it remains undefined for calibration-free LIBS (CF-LIBS). In this work, a statistical LOD definition is proposed for CF-LIBS based on the density of emitters, plasma parameters, and measurement uncertainties. The emission line intensities are assumed to have uncertainties that can be estimated by the standard deviation of the background signal. The LOD proposed in this work is determined through the propagation of uncertainties applied to CF-LIBS calculations, and is shown to depend on plasma temperature and emission line properties. It was validated using carbon and calcium in sodium chloride samples, yielding values comparable to those obtained with the LOD definition for calibration curves. In addition to allowing for LOD estimation in CF-LIBS, the proposed expression also provides a sound and thorough LOD definition for LIBS quantitative analyses.