## Abstract

Laser-induced breakdown spectroscopy (LIBS) coupled with the linear multivariate regression method was utilized to analyze chromium (Cr) quantitatively in potatoes. The plasma was generated using a Nd:YAG laser, and the spectra were acquired by an Andor spectrometer integrated with an ICCD detector. The models between intensity of LIBS characteristic line(s) and concentration of Cr were constructed to predict quantitatively the content of target. The unary, binary, ternary, and quaternary variables were chosen for verifying the accuracy of linear regression calibration curves. The intensity of characteristic lines Cr (CrI: 425.43, 427.48, 428.97 nm) and Ca (CaI: 422.67, 428.30, 430.25, 430.77, 431.86 nm) were used as input data for the multivariate calculations. According to the results of linear regression, the model of quaternary linear regression was established better in comparing with the other three models. A good agreement was observed between the actual content provided by atomic absorption spectrometry and the predicted value obtained by the quaternary linear regression model. And the relative error was below 5.5% for validation samples S1 and S2. The result showed that the multivariate approach can obtain better predicted accuracy than the univariate ones. The result also suggested that the LIBS technique coupled with the linear multivariate calibration method could be a great tool to predict heavy metals in farm products in a rapid manner even though samples have similar elemental compositions.

© 2015 Optical Society of America

Full Article | PDF Article**OSA Recommended Articles**

Baohua Zhang, Wen Sha, Yongcheng Jiang, and Zhifeng Cui

Appl. Opt. **58**(12) 3277-3281 (2019)

Mingyin Yao, Gangfu Rao, Lin Huang, Muhua Liu, Hui Yang, Jinyin Chen, and Tianbing Chen

Appl. Opt. **56**(29) 8148-8153 (2017)

Lin Huang, Tianbing Chen, Xiuwen He, Hui Yang, Caihong Wang, Muhua Liu, and Mingyin Yao

Appl. Opt. **56**(1) 24-28 (2017)