Zhao Jin-Hui, Yuan Hai-Chao, Liu Mu-Hua, Xiao Hai-Bin, Hong Qian
Guang Pu Xue Yu Guang Pu Fen Xi. 2014 Apr;34(4):1012-6.
In order to realize the rapid determination of ractopamine content in pork, quantitative determination model of ractopamine content in pork was established by using three-dimensional synchronous fluorescence spectrum coupled with alternating penalty trilinear decomposition (APTLD). Firstly, the generation mechanism of the fluorescence spectrum for ractopamine and three-dimensional synchronous fluorescence spectrum for samples were analyzed. Secondly, concentration quenching phenomenon of fluorescence of ractopamine in pork extract was investigated. Thirdly, the number of components for three linear decomposition of APTLD was set as 2 by using the core consistency diagnostic method, and the calibration curve of the relative fluorescence intensity of ractopamine between pork extract and the training sample was established for the correction of relative fluorescence intensity of prediction samples. Finally, three-dimensional synchronous fluorescence spectrum combined with APTLD was used to build the prediction model of ractopamine content in pork. The experimental results showed that the method adopted in the paper could better solve the problem of serious synchronous fluorescence spectrum overlapping between ractopamine in pork samples and backgrounds, and leave out some trivial process of chemical separation for the identification of ractopamine in pork. The determination coefficient (R2) and the root mean squared error of prediction (RMSEP) for the model proposed in this paper were 0.986 3 and 0.496 6 mg x L(-1), respectively. The method in this paper has achieved the goal of rapid quantitative detection of ractopamine content in pork.
为实现猪肉中莱克多巴胺含量的快速测定，采用三维同步荧光光谱结合交替惩罚三线性分解法（APTLD）建立了猪肉中莱克多巴胺含量的定量测定模型。首先，分析了莱克多巴胺荧光光谱的产生机理及样品的三维同步荧光光谱。其次，研究了猪肉提取物中莱克多巴胺荧光的浓度猝灭现象。再次，采用核心一致性诊断法将APTLD三线性分解的组分数设为2，并建立了猪肉提取物中莱克多巴胺相对荧光强度与训练样本间的校准曲线，用于预测样本相对荧光强度的校正。最后，利用三维同步荧光光谱结合APTLD建立了猪肉中莱克多巴胺含量的预测模型。实验结果表明，本文采用的方法能较好地解决猪肉样品中莱克多巴胺与背景同步荧光光谱严重重叠的问题，省去了猪肉中莱克多巴胺鉴定时繁琐的化学分离过程。本文所建模型的测定系数（R2）和预测均方根误差（RMSEP）分别为0.986 3和0.496 6 mg·L-1。本文方法实现了猪肉中莱克多巴胺含量的快速定量检测目标。
