Automatic correction of continuum background in Laser-induced Breakdown Spectroscopy using a model-free algorithm

Abstract We report on the first use of a model-free algorithm (designed to remove baseline in NMR spectra) for background correction of LIBS data. We also present our updated version of the algorithm that is potentially a better candidate for use in automated data processing routines in LIBS. This updated algorithm requires no assumptions to be made on the character of the background or type of spectra, uses no thresholds, and works equally well for high and low signal-to-noise LIBS spectra. The performance of the algorithm was evaluated using a range of randomly generated synthetic spectra and experimental LIBS spectra of varying complexity and noise characteristics. The analytical performance of the algorithm was evaluated by comparing the accuracy of LIBS measurements achieved when using untreated background, when subtracting the background using the original Friedrichs's algorithm and when subtracting the background using our updated algorithm.