【作者】 陈健；

【导师】 赵进辉； 刘木华； 王华中；

【作者基本信息】 江西农业大学 ， 农业硕士（专业学位）， 2020， 硕士

【摘要】 禽肉是一种具有很高营养价值的食品,也是生活饮食中不可缺少的食品之一。在家禽养殖中,为了防治家禽的疾病,不可避免的会使用到抗生素。然而,部分抗生素不易降解,不合理的使用会导致其残留在家禽体内,进而威胁到人们的身体健康。为了确保禽肉产品的质量安全,一些国内外学者一直在探索研究抗生素残留快速检测技术。传统的检测抗生素残留方法前处理复杂、耗时长、成本高,较难实现禽肉中抗生素残留的快速检测。本文以禽肉(鸡肉、鸭肉)为载体,以磺胺二甲基嘧啶(sulfamethazine,SM2)、氧氟沙星(ofloxacin,OFL)、甲磺酸达氟沙星(danofloxacin mesylate,DFM)、盐酸沙拉沙星(sarafloxacin hydrochloride,SARH)和盐酸强力霉素(doxycycline hydrochloride,DCH)为研究对象,采用同步荧光技术结合化学计量学方法探索禽肉中抗生素残留的快速检测方法。主要的研究工作如下:(1)采用同步荧光技术结合化学计量学方法实现了禽肉中SM2和OFL残留的快速检测。首先,分析了SM2标准溶液、OFL标准溶液、空白禽肉提取液和含SM2和OFL的禽肉提取液的三维同步荧光光谱,确定了禽肉中SM2和OFL残留检测的波长差(Δλ)分别为150 nm和210 nm,荧光激发峰分别为292.5 nm和295 nm。其次,采用单因素试验考察了β-巯基乙醇和邻苯二甲醛溶液加入量及时间对荧光强度的影响,确定了鸡肉中SM2和OFL残留的最佳检测条件为:β-巯基乙醇溶液300μL、邻苯二甲醛溶液25μL和采集时间44 min;确定了鸭肉中SM2和OFL残留的最佳检测条件为:β-巯基乙醇溶液400μL、邻苯二甲醛溶液25μL和采集时间40 min。最后,利用峰高法和峰面积法分别建立了鸡肉中SM2和OFL残留的预测模型。试验结果表明,与基于峰面积法的预测模型相比,基于峰高法的预测模型的综合评价更好。基于峰高法的SM2残留预测模型的线性方程为y=0.550 5x+18.884,训练集决定系数(coefficient of determination for the training set,R_T~2)为0.919 4,预测集决定系数(coefficient of determination for the prediction set,R_P~2)为0.897 3,预测集均方根误差(root mean square error for the prediction set,RMSEP)为6.060 5 mg/kg,回收率处于76.1～115.2%之间,相对标准偏差(relative standard deviation,RSD)处于2.7～7.0%之间。基于峰高法的OFL残留预测模型的线性方程为y=7.783 8x+15.919,R~2_T为0.973 8,R~2_P为0.997 3,RMSEP为0.539 2 mg/kg,回收率处于96.7～110.1%之间,RSD处于2.8～10.0%之间。此外,利用峰高法和峰面积法分别建立了鸭肉中SM2和OFL残留的预测模型。试验结果表明,与基于峰面积法的预测模型相比,基于峰高法的预测模型的综合评价更好。基于峰高法的SM2残留预测模型的线性方程为y=0.501 7x+19.049,R~2_T为0.8447,R~2_P为0.815 6,RMSEP为7.950 9 mg/kg,回收率处于81.7～155.1%之间,RSD处于4.1～6.7%之间。基于峰高法的OFL残留预测模型的线性方程为y=7.320 6x+10.705,R~2_T为0.996 6,R~2_P为0.996 2,RMSEP为0.526 7 mg/kg,回收率处于96.4～111.2%之间,RSD处于2.9～6.8%之间。(2)采用同步荧光技术结合化学计量学方法实现了禽肉中DFM和OFL残留的快速检测。首先,分析了DFM标准溶液、OFL标准溶液、空白禽肉提取液和含DFM和OFL的禽肉提取液的同步荧光光谱,确定了禽肉中DFM和OFL残留的检测Δλ分别为130 nm和200 nm,荧光激发峰分别为288 nm和325 nm。其次,采用单因素试验考察了氢氧化钠溶液浓度和表面活性剂种类对荧光强度的影响,确定了鸡肉中DFM和OFL残留的最佳检测条件为:氢氧化钠溶液浓度0.1mol/L和SDS溶液浓度0.1 mol/L;确定了鸭肉中DFM和OFL残留的最佳检测条件为:氢氧化钠溶液浓度0.1 mol/L和SDS溶液浓度0.1 mol/L。最后,利用线性回归、偏最小二乘回归(partial least squares regression,PLSR)和多元线性回归(multiple linear regression,MLR)算法分别建立了鸡肉中DFM和OFL残留的预测模型。试验结果表明,与基于线性回归和MLR的DFM残留预测模型相比,基于PLSR的DFM残留预测模型的综合评价更好,其R~2_P为0.978 3,RMSEP为1.934 2 mg/kg,相对预测误差(ratio of prediction to deviation,RPD)为5.876 5;与基于线性回归和PLSR的OFL残留预测模型相比,基于MLR的OFL残留预测模型的综合评价更好,其R~2_P为0.895 0,RMSEP为3.859 8 mg/kg,RPD为2.509 1。此外,利用线性回归、PLSR和MLR算法分别建立了鸭肉中DFM和OFL残留的预测模型。试验结果表明,与基于PLSR和MLR的预测模型相比,基于线性回归的预测模型的综合评价更好。基于线性回归的DFM残留预测模型的R~2_P为0.968 2,RMSEP为1.619 2 mg/kg,RPD为4.711 7。基于线性回归的OFL残留预测模型的R_P~2为0.964 3,RMSEP为2.244 0 mg/kg,RPD为4.751 3。(3)采用同步荧光技术结合化学计量学方法实现了鸡肉中SARH和DCH残留的快速检测。首先,分析了SARH标准溶液、DCH标准溶液、空白鸡肉提取液和含SARH和DCH的鸡肉提取液的三维同步荧光光谱,确定了鸡肉中SARH和DCH残留检测的Δλ都为110 nm,荧光激发峰分别为320 nm和381 nm。其次,采用单因素试验考察了硫酸镁溶液浓度和SDS溶液浓度及时间对荧光强度的影响,确定了鸡肉中SARH和DCH残留的最佳检测条件为:硫酸镁溶液浓度0.375 mol/L、SDS溶液浓度0.30 mol/L和采集时间12 min。最后,利用MLR、PLSR和支持向量回归(support vector regression,SVR)算法分别建立了鸡肉中SARH和DCH残留的预测模型。试验结果表明,与基于MLR和SVR的预测模型相比,基于PLSR的预测模型的综合评价更好。基于PLSR的SARH残留预测模型的R~2_P为0.846 5,RMSEP为0.344 1 mg/kg,RPD为2.588 2。基于PLSR的DCH残留预测模型的R~2_P为0.914 1,RMSEP为5.890 9 mg/kg,RPD为3.243 5。本文采用同步荧光技术结合化学计量学方法探索了禽肉中SM2、OFL、DFM、SARH和DCH残留的快速检测方法,在此基础上可以进一步优化和拓宽同步荧光技术在禽肉中其它抗生素残留检测中的应用。更多还原

【Abstract】 Poultry is a kind of food with high nutritional value,and it is also one of the indispensable foods in daily life.In poultry farming,antibiotics are inevitably used in order to prevent and treat various diseases of poultry.However,some antibiotics are not easily degraded,and unreasonable use will cause them to remain in poultry,which threatens people’s health.To ensure the quality and safety of poultry products,some scholars at home and abroad have been exploring and researching rapid detection technologies for antibiotic residues.Traditional methods for detecting antibiotic residues require complex preprocessing,long time and high cost,making it difficult to achieve rapid detection of antibiotic residues in poultry.In this work,poultry(chicken,duck meat)was used as a carrier,sulfamethazine(SM2),ofloxacin(OFL),danofloxacin mesylate(DFM),sarafloxacin hydrochloride(SARH)and doxycycline hydrochloride(DCH)were used as research objects.Synchronous fluorescence technology coupled with chemometric methods were used to explore a rapid detection method for antibiotic residues in poultry.The main research work is as follows:(1)The rapid detection for SM2 and OFL residues in poultry was achieved through synchronous fluorescence technology coupled with chemometric methods.First of all,three-dimensional synchronous fluorescence spectra of SM2 standard solution,OFL standard solution,poultry extract without antibiotics and poultry extract containing SM2and OFL were analyzed,and the wavelength difference(Δλ)of SM2 and OFL were respectively determined as 150 nm and 210 nm,and the fluorescence excitation peaks of SM2 and OFL were respectively determined as 292.5 nm and 295 nm for the detection of SM2 and OFL residues in poultry.Subsequently,the effects of the addition amounts ofβ-mercaptoethanol solution and o-phthalaldehyde solution,as well as time,on the fluorescence intensities were investigated through the single factor test.The best detection conditions of SM2 and OFL residues in chicken were as follows:β-mercaptoethanol solution of 300μL,o-phthalaldehyde solution of 25μL and time of 44 min.Besides,the best detection conditions of SM2 and OFL residues in duck meat were as follows:β-mercaptoethanol solution of 400μL,o-phthalaldehyde solution of 25μL and time of 40 min.Finally,the prediction models of SM2 and OFL residues in chicken were established through the algorithms of the peak height and peak area,respectively.The experimental results showed that the prediction model based on the peak height algorithm has the better comprehensive evaluation by comparison of the prediction model based on the peak area algorithm.For the prediction model of SM2 residues based on the peak height algorithm,the linear equation was y=0.550 5x+18.884.The coefficient of determination for the training set(R_T~2)and prediction set(R_P~2)were 0.919 4 and 0.897 3,respectively.The root mean square error for the prediction set(RMSEP)was 6.060 5 mg/kg,and the recovery was in the range of 76.1～115.2%,and the relative standard deviation(RSD)was in the range of 2.7～7.0%.For the prediction model of OFL residues based on the peak height algorithm,the linear equation was y=7.783 8x+15.919.The R~2_Tand R~2_Pwere 0.973 8 and0.997 3,respectively.The RMSEP was 0.539 2 mg/kg.The recovery and RSD were in the range of 96.7～110.1%and 2.8～10.0%,respectively.Besides,the prediction models of SM2 and OFL residues in duck meat were established through the algorithms of the peak height and peak area,respectively.The experimental results show that the comprehensive evaluation of the prediction model based on the peak height algorithm better than the prediction model based on the peak area algorithm.For the prediction model of SM2 residues based on the peak height algorithm,the linear equation was y=0.501 7x+19.049.The R~2_Tand R~2_Pwere 0.844 7 and 0.815 6,respectively.The RMSEP was 7.950 9 mg/kg.The recovery and RSD were in the range of81.7～155.1%and 4.1～6.7%,respectively.For the prediction model of OFL residues based on the peak height algorithm,the linear equation was y=7.320 6x+10.705.The R~2_Tand R_P~2were 0.996 6 and 0.996 2,respectively.The RMSEP was 0.526 7 mg/kg.The recovery and RSD were in the range of 96.4～111.2%and 2.9～6.8%,respectively.(2)The rapid detection for DFM and OFL residues in poultry was achieved through synchronous fluorescence technology coupled with chemometric methods.First of all,the synchronous fluorescence spectra of DFM standard solution,OFL standard solution,poultry extract without antibiotics and poultry extract containing DFM and OFL were analyzed,and theΔλof DFM and OFL were respectively determined as 130 nm and 200nm,and the fluorescence excitation peaks of DFM and OFL were respectively determined as 288 nm and 325 nm for the detection of DFM and OFL in poultry,respectively.Subsequently,the effects of the concentrations of sodium hydroxide solution and the type of surfactant on the fluorescence intensities were investigated through the single factor test.The best detection conditions of DFM and OFL residues in chicken were as follows:the concentration of sodium hydroxide solution of 0.1 mol/L,and the concentration of SDS solution of 0.1 mol/L.The best detection conditions of DFM and OFL residues in duck meat were as follows:the concentration of sodium hydroxide solution of 0.1 mol/L,and the concentration of SDS solution of 0.1 mol/L.Finally,the prediction models of DFM and OFL residues in chicken were established through the algorithms of the linear regression,partial least squares regression(PLSR),and multiple linear regression(MLR),respectively.The experimental results showed that the comprehensive evaluation of the prediction model of DFM residues based on the PLSR algorithm was best among these algorithms.The R~2_Pand RMSEP were 0.978 3 and 1.934 2mg/kg,respectively.The ratio of prediction to deviation(RPD)was 5.876 5.The comprehensive evaluation of the prediction model of OFL residues based on the MLR algorithm was best among these algorithms.The R_P~2,RMSEP,and RPD were 0.895 0,3.859 8 mg/kg,and 2.509 1,respectively.Besides,the prediction models of DFM and OFL residues in duck meat were established through the algorithms of the linear regression,PLSR,and MLR,respectively.The experimental results showed that the comprehensive evaluation of the prediction model based on the linear regression algorithm was best among these algorithms.For the prediction model of DFM residues based on the linear regression algorithm,the R_P~2,RMSEP,and RPD were 0.968 2,1.619 2 mg/kg,and 4.711 7,respectively.For the prediction model of OFL residues based on the linear regression algorithm,the R_P~2,RMSEP,and RPD were 0.964 3,2.244 0 mg/kg,and 4.751 3,respectively.(3)The rapid detection for SARH and DCH residues in chicken was achieved through synchronous fluorescence technology coupled with chemometric methods.First of all,three-dimensional synchronous fluorescence spectra of SARH standard solution,DCH standard solution,chicken extract without antibiotics and chicken extract containing SARH and DCH were analyzed,and theΔλof both SARH and DCH were determined as 110 nm,and the fluorescence excitation peaks of SARH and DCH were respectively determined as320 nm and 381 nm for the detection of SARH and DCH residues in chicken.Subsequently,the effects of the concentrations of magnesium sulfate solution and SDS solution,as well as time,on the fluorescence intensities were investigated through the single factor test.The best detection conditions of SARH and DCH residues in chicken were as follows:the concentration of magnesium sulfate solution of 0.375 mol/L,the concentration of SDS solution of 0.30 mol/L and time of 12 min.Finally,the prediction models of SARH and DCH residues in chicken were established through the algorithms of the MLR,PLSR,and support vector regression(SVR),respectively.The experimental results showed that the comprehensive evaluation of the prediction model based on the PLSR algorithm was best among these algorithms.For the prediction model of SARH residues based on the PLSR algorithm,the R_P~2,RMSEP,and RPD were 0.846 5,0.344 1 mg/kg,and 2.588 2,respectively.For the prediction model of DCH prediction based on the PLSR algorithm,the R_P~2,RMSEP,and RPD were 0.914 1,5.890 9 mg/kg,and 3.243 5,respectively.In this thesis,a method for the rapid detection of SM2,OFL,DFM,SARH,and DCH residues in poultry was explored using synchronous fluorescence technology combined with chemometric methods.Based on this,the application of synchronous fluorescence technology can be further optimized and expanded for the detection of other antibiotic residues in poultry.更多还原