Abstract：Heavy metals pollution in foodstuffs is more and more serious. It is impossible to satisfy the modern agricultural development by conventional chemical analysis. Laser induced breakdown spectroscopy (LIBS) is an emerging technology with the characteristic of rapid and nondestructive detection. But LIBS’s repeatability, sensitivity and accuracy has much room to improve. In this work, heavy metal Cu in Gannan Navel Orange which is the Jiangxi specialty fruit will be predicted by LIBS. Firstly, the navel orange samples were contaminated in our lab. The spectra of samples were collected by irradiating the peel by optimized LIBS parameters. The laser energy was set as 20 mJ, delay time of Spectral Data Gathering was set as 1.2 μs, the integration time of Spectral data gathering was set as 2 ms. The real concentration in samples was obtained by AAS (atom absorption spectroscopy). The characteristic variables Cu Ⅰ 324.7 and Cu Ⅰ 327.4 were extracted. And the calibration model was constructed between LIBS spectra and real concentration about Cu. The results show that relative error of the predicted concentrations of three relational model were 7.01% or less, reached a minimum of 0.02%, 0.01% and 0.02% respectively. The average relative errors were 2.33%, 3.10% and 26.3%. Tests showed that different characteristic variables decided different accuracy. It is very important to choose suitable characteristic variable. At the same time, this work is helpful to explore the distribution of heavy metals between pulp and peel.
Key words：LIBS;Navel orange peel;Cu;Extraction of characteristic variable
黎文兵，姚明印*，黄 林，陈添兵，郑建鸿，樊十全，刘木华，何秀文，林金龙，欧阳静怡 . 特征变量提取方法对脐橙皮中Cu元素LIBS检测精度的影响 [J]. 光谱学与光谱分析, 2015, 35(07): 2021-2024.
LI Wen-bing, YAO Ming-yin*, HUANG Lin, CHEN Tian-bing, ZHENG Jian-hong, FAN Shi-quan, LIU Mu-hua, HE Xiu-wen, LIN Jin-long, OUYANG Jing-yi. Effect of Characteristic Variable Extraction on Accuracy of Cu in Navel Orange Peel by LIBS . SPECTROSCOPY AND SPECTRAL ANALYSIS, 2015, 35(07): 2021-2024.
