二维相关近红外光谱检测牛奶中的三聚氰胺

配置合格的纯牛奶样本及含有三聚氰胺质量浓度范围为0.01g/L～3g/L的掺杂牛奶样本各20个,并采集其近红外光谱。以牛奶中掺杂三聚氰胺浓度为外扰,构建二维相关同步谱,研究其相关谱特性。在此基础上,结合偏最小二乘判别分析法(PLS-DA)建立定性模型,可以实现纯牛奶与掺伪牛奶的定性鉴别,正确识别率达100%。同时,将二维相关近红外同步谱矩阵与偏最小二乘法(PLS)结合起来,建立定量分析牛奶中掺杂三聚氰胺的数学模型。对未知样品的预测相关系数R达到0.98,预测均方根误差(RM-SEP)为0.18g/L,说明基于同步相关谱矩阵建立定量分析的数学模型是可行的。该方法无需样品处理,成本低,为快速检测掺伪牛奶提供了一种新的途径。

Detection of Melamine in Milk by Two-dimensional Near-infrared Correlation Spectroscopy

20 pure milk samples and 20 adulterated milk samples with different mass concentration of melamine(0.01~3 g/L) were prepared and their near infrared spectra were collected.The two-dimensional correlation spectra were calculated under the perturbation of adulteration concentration.The characteristics of two-dimensional correlation spectra of adulterated milk were studied.Partial least squares discriminate analysis(PLS-DA) was applied to differentiate the pure milk samples and the adulterated milk samples.The classification accuracy was 100%.At the same time,the model for quantitative analysis of melamine concentration in the adulterated milk samples was constructed by partial least square(PLS) combined with two-dimensional near infrared correction spectra.Results showed that the root mean square error of prediction(RMSEP) was 0.18 g/L.The correlation coefficient between reference values and predicted values was 0.98.The method does not rely on sample separation,and provides alternative way to detect the adulterant of milk.