Improving LIBS analysis of non-flat heterogeneous samples by signals mapping.

Heterogeneous material analysis by the laser-induced breakdown spectroscopy (LIBS) technique is challenging in real practice due to requirements for representative sampling and non-flat surfaces of the samples. Methods complementary to LIBS (plasma imaging, plasma acoustics, sample surface color imaging) have been introduced to improve zinc (Zn) determination in soybean grist material by LIBS. The detailed statistical study revealed that atomic/ionic lines emission and other LIBS signals were distributed normally except for acoustics signals. The correlation between LIBS and complementary signals was rather poor due to the large variability of the particle properties of soybean grist material. Still, analyte line normalization on plasma background emission was rather simple and effective for Zn analysis but required a few hundred spot samplings for representative Zn quantification. Non-flat heterogeneous samples (soybean grist pellets) were analyzed by LIBS mapping but it was demonstrated that the choice of sampling area is crucial for reliably analyte determination.