基于SERS与3D-SFS的禽肉中兽药多残留的快速同时检测方法研究

The studying team discovered that the fluorescent spectroscopy was difficult to meet the rapid and simultaneous detection demands of different types of multi-antibiotic residues in duck meat according to the research of the last National Natural Science Foundation of China. Therefore, spectral characteristics of surface-enhanced Raman spectroscopy (SERS) and three-dimensional synchronous fluorescence spectroscopy (3D-SFS) for veterinary drug residues in poultry (chicken, duck meat) are investigated based on the comprehensive application of surface-enhanced Raman spectroscopy, three-dimensional synchronous fluorescence spectroscopy, multi-sensor data fusion and etc. in order to find a rapid detection method of veterinary drug multi-residues (antibiotics and hormones) in poultry..Firstly, the spectral information of SERS (silver colloid, gold colloid and core-shell as nano-reinforcement substrates) and 3D-SFS of typical veterinary drug residues in poultry is obtained. Secondly, the chemical methods are applied to determine the actual contents of veterinary drug residues in poultry. Thirdly, the characteristic parameters are extracted and selected using three-dimensional matrix second-order calibration algorithm, wavelet analysis, particle swarm optimization algorithm and etc. Lastly, the chemometrics methods，such as support vector machine, relevant vector machine and multi-sensor data fusion, are applied to establish the detection models of veterinary drug multi-residues in poultry..The completion of this project will acquire a rapid detection method of veterinary drug multi-residues in poultry with technical innovation, establish the corresponding rapid detection system, realize the automatic detection of veterinary drug multi-residues in poultry, and has the important significance for the edible and safe quality detection of poultry.

课题组在前一个国家基金研究中发现，单一的荧光光谱技术较难满足鸭肉中多类多种抗生素残留的快速同时检测要求。基于此，我们拟综合运用表面增强拉曼光谱(SERS)、三维同步荧光光谱（3D-SFS）和多传感器数据融合等方法来分析禽肉（鸡肉、鸭肉）中兽药残留的SERS与3D-SFS光谱特性，以期找到一种禽肉中兽药(抗生素类与激素类)多残留的快速同时检测方法。.项目先获取禽肉中典型兽药残留的SERS（以金胶、银胶、核壳作为纳米增强基底）与3D-SFS光谱信息；然后用化学方法测定其真实残留含量；运用三维数阵二阶校正算法、小波分析、粒子群算法等方法进行特征参数的提取与选择；进一步用支持向量机、相关向量机、多传感器数据融合等方法建立禽肉中兽药多残留检测模型。.项目的完成将获得一种具有技术创新的禽肉中兽药多残留的快速同时检测方法，并建立相应的快速检测系统，实现其检测自动化，对禽肉食用安全品质检测具有重要意义。

本项目以鸡肉、鸭肉作为研究对象，结合表面增强拉曼技术、同步荧光光谱技术和模式识别等技术知识，进行禽肉中兽药多残留的快速同时检测方法研究。主要研究内容和重要研究结果如下：应用PCA结合SVM分别建立了鸡肉和鸭肉中磺胺吡啶和磺胺二甲嘧啶残留检测的SERS鉴别模型。应用PCA结合LDA分别建立了鸡肉和鸭肉中新霉素和氯霉素残留检测的SERS模型。分别建立了鸡肉和鸭肉中呋喃它酮代谢物残留检测的SERS预测模型，这两种预测模型的决定系数分别为0.9245和0.9961。分别建立了鸡肉和鸭肉中呋喃妥因代谢物残留检测的SERS预测模型，这两种预测模型的决定系数分别为0.9955和0.9976。利用主成分-线性判别分析法分别建立了鸡肉和鸭肉中呋喃它酮代谢物和呋喃妥因代谢物残留检测的SERS鉴别模型。分别建立了鸭肉中丙酸睾酮、诺龙和己烯雌酚残留检测的SERS预测模型，其模型的决定系数分别为0.9761、0.9436和0.9424。应用PCA结合SVM建立了鸡肉中的丙酸睾酮和诺龙残留检测的SERS识别模型。应用PCA结合LDA建立了鸡肉中盐酸强力霉素和泰乐菌素残留检测的的SERS预测模型。建立了基于PCA-SVM的鸭肉中盐酸强力霉素和泰乐菌素残留检测的的SERS预测模型。建立了基于峰高法的鸡肉和鸭肉中磺胺二甲基嘧啶和氧氟沙星残留检测的同步荧光光谱预测模型。建立了鸡肉和鸭肉中甲磺酸达氟沙星和氧氟沙星残留检测的同步荧光光谱预测模型。建立了基于PLSR的鸡肉中盐酸沙拉沙星和盐酸强力霉素残留检测的同步荧光光谱预测模型。以上研究成果对禽肉食用安全品质检测的实现具有重要意义。