Minimizing matrix effect for quantitative analyses of soil samples using the LIBS technique

Abstract

Laser-induced breakdown spectroscopy (LIBS) is a promising alternative for the quantification of nutrients and contaminants in soil samples because it requires little sample preparation, is fast, and does not generate harmful residues to the environment. Despite these advantages, LIBS technique has not been established for routine analysis due to low reproducibility, especially because of matrix effects, and low sensitivity for some elements, e.g., Pb, S, and P. In this work, we propose: i) the development of quantitative models based on the calibration-free LIBS (CF-LIBS) and the one-point calibration (OPC) method to minimize matrix effects for soil samples and ii) protocols of sample preparation to improve the sensitivity of the technique. Originally developed in 1999 to minimize matrix effects, CF-LIBS is still barely used today due to its difficulty in use and lack of calibration of emitter densities in the plasma. Proposed in 2013, OPC is a method that empirically corrects the intensity of the emission lines of the elements and can be used to calibrate the density of emitters in the plasma. To increase the sensitivity of the LIBS technique, we will test the addition of some compounds that can intensify the emission lines of the elements of interest. In this proposal, we will use soil samples with varying characteristics of chemical composition and texture. We will also use synthetic samples when it is convenient to test models and methods under controlled and known conditions. Analysis of the results of the proposed models and methods will be done by the coefficient of determination and the root-mean-square error of prediction. The sensitivity of the technique will be evaluated by the limit of detection and the signal-to-noise ratio.(AU)