An echellograms-based quantitative analysis method of laser-induced breakdown spectroscopy.

The high-resolution broadband LIBS spectrum is typically acquired using an echelle spectrometer and array detectors. However, the full frame (i.e. , echellogram) captured by detector is commonly converted to 1D spectrum for the further analysis, which would both limit the detection speed of LIBS and potentially degenerate the analytical performance. In this work, we propose an echellograms-based analysis method for the quantification of minor metal elements in steel. Two metal elements, Mn and Ti, are chosen to validate the feasibility of the analytical method. The LASSO algorithm is used to build the multivariate regression model using LIBS spectra and echellograms respectively. The analytical performance obtained with the echellograms-based model is compared with that of spectra-based model. The efficiency of the echellograms-based quantitative analysis method is demonstrated with better prediction accuracy and precision. Notably, the echellogram-based regression model involves only tens to thousands of pixels from the echellogram, which can be selectively read out using a CMOS detector. The combination of echellogram-based quantitative analysis method and CMOS detector provides new ideas for high-frequency LIBS detection. [Display omitted] • Accurate determination of minor elements in steel using echellograms-based analysis method of LIBS. • Comparison of analytical performance between spectra-based regression models and echellograms-based regression models. • Observing the influence of echellogram's size on analytical performance. • Analyzing the pixels involved in regression model and corresponding positions in echellogram. [ABSTRACT FROM AUTHOR]