Laser-Induced Breakdown Spectroscopy; Multivariate classification; Echelle spectrometer; Czerny-Turner spectrometer;

The objective of this work was to assess a part of echelle Laser-Induced Plasma spectra (ranging from 200 to 1000 nm) that could be most effectively employed for rocks classification. Therefore, a 60 nm wide spectral window mask was iteratively moved over the broadband echelle spectra. Each created narrow artificial spectral windows (60 nm) was used for the classification of rock samples using various Multivariate Data Analysis (MVDA) algorithms, reaching more than 99% of the overall accuracy in certain cases. Afterwards, the Czerny-Turner spectrometer (having higher sensitivity compared to the echelle spectrometer) was aligned to the a priori selected and the most effective spectral regions and rocks samples were re-measured. Consequently the MVDA analyses were utilized again, providing also satisfying classification results yielding more than 99% of the overall accuracy. Measurements of 28 sedimentary ores (certified reference materials) were done utilizing commercially available X-Trace device (AtomTrace), where spectrometers in both configurations (echelle and Czerny-Turner) were exploited.