Multivariate Analysis of Laser-Induced Breakdown Spectroscopy Spectra of Soil Samples

• DOI: 10.1097/ss.0b013e3181f516ea
• Publication Date: 2010-09-03T19:23:45Z
• Authors: Ningfang Yang; Neal S. Eash; Jaehoon Lee; Madhavi Z. Martin; Yong-Seon Zhang; Forbes R. Walker; Jae E. Yang
• Keywords: Univariate; Laser-induced breakdown spectroscopy; Soil test

Laser-induced breakdown spectroscopy (LIBS) is a rapid quantitative analytical technique that can be used to determine the elemental composition of numerous sample matrices, and it has been successfully applied in many types samples. However, for chemically physically complex soil samples, its ability controversial. Multivariate techniques have great potential analyzing LIBS spectra. To demonstrate feasibility as an alternative quantitatively analyze univariate partial least square (PLS) are spectra 12 samples build calibration models predicting Cu Zn concentrations. The results show PLS significantly improve compared with technique. normalized root mean error (NRMSE) r2 16.60% 0.71 18.80% 0.62 prediction 18.97% 22.81% 0.45 Zn. For using spectral range 300 350 nm, NRMSE 1.94% 0.99 both 7.90% 0.94 8.14% prediction, respectively. Compared technique, improves 87.53% 87.78% 44.47% 53.44% Zn, indicate analysis.