Open AccessQuantitative analysis of laser induced breakdown spectroscopy of Pb in navel orange based on multivariate calibration
Author(s) -
Tianbing Chen,
Mingyin Yao,
Muhua Liu,
Lin Yongzeng,
Wenbing Li,
Zheng Mei-Lan,
Huamao Zhou
Publication year - 2014
Publication title -
wuli xuebao
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.199
H-Index - 47
ISSN - 1000-3290
DOI - 10.7498/aps.63.104213
Subject(s) - laser induced breakdown spectroscopy , multivariate statistics , linear regression , analytical chemistry (journal) , bayesian multivariate linear regression , spectroscopy , calibration , navel orange , materials science , statistics , chemistry , mathematics , chromatography , physics , quantum mechanics , horticulture , biology
The detection accuracy of laser induced breakdown spectroscopy (LIBS) is affected by system parameters, ambient gas, matrix effect, sample morphology, calibration methods etc. Heavy metals in Gannan navel orange are determined by LIBS in our laboratory. The experimental parameters are optimized. In this work, multivariate linear regression model is used to predict the concentration of Pb element in navel oranges. The real concentration of Pb is quantitatively determined by atomic absorb spectroscopy (AAS). The concentration is set as dependent variable, while the intensity of Pb I 405.78 nm, the intensity sum of Ca Ⅱ 393.37 nm and Ca Ⅱ 396.84 nm, and the integrated intensity in a range of 405.03-405.96 nm are taken as independent variable. The calibration results indicate that the maximum relative error between the predicted Pb concentration from the multiple linear regression model and the measured one by the AAS is 12.99%, and the average relative error of the samples is 4.87%. And the fitting degree of the results of two methods is 0.995. The result shows that the multivariate calibration method can utilize the information about the spectra and reduce the influence of the matrix effect. The multivariate linear regression model is proved to be feasible in improving the prediction accuracy of LIBS.