Statistical evaluation of spectral interferences in laser-induced breakdown spectroscopy

Abstract

Line broadening due to plasma processes or the instrument degrade spectral resolution leading to uncertainty in the elemental profile obtained by optical emission spectroscopy. A novel approach to quantify spectral interferences in laser-induced breakdown spectroscopy (LIBS) is introduced. This algorithm establishes a statistical interference factor (SIF) to quantify spectral interferences for individual line assignment in LIBS while using the full information provided by a spectrum. This factor combines fundamentals of plasma emission with a Bayesian analysis of the experimental spectrum using eventual prior knowledge about the sample and/or the conditions of analysis. Two types of prior knowledge were used to interpret spectra from a pure silicon sample as well as a NIST SRM 610 glass sample and alumina with a controlled Ni contamination at different concentrations. Knowledge of the elemental composition confirmed the existence of spectral lines with high sensitivity to interference while other lines are more appropriate for analysis due to their consistent SIFs despite having knowledge of the composition.