牡蛎中氨基酸的傅里叶变换近红外光谱法快速定量测定

本文通过乘法散射校正(multiplicative scattering correction MSC)、一阶求导(1st Derivative,Db1)和Norris平滑等傅里叶变换近红外光谱定量分析数学模型,用交叉验证法确定了最优化建模参数,构建了牡蛎中氨基酸的快速定量测定方法。结果表明,牡蛎近红外光谱与氨基酸含量化学计量学拟合,线性相关系数为0.9860,交叉验证后线性相关系数为0.9794,对牡蛎样品进行检测,结果与传统的氨基酸分析仪的测试结果相比,相对误差小于2.5%,方法的精密度(RSD)小于1.0%,本文建立的傅里叶变换近红外光谱法定量分析牡蛎中氨基酸具有快速、准确、操作简便等特点,测定重复性优于氨基酸分析仪法。

Quantitative analysis of amino acids in oyster by FT-NIR spectroscopy

In this study, spectra pretreatment methods of multiplicative scattering correction (MSC), Norris smoothing, and 1st Derivative (Db1) were applied to develop a FT-NIR mathematics model for quantitative analysis of amino acids in oyster. Optimum parameters of the model were studied using a cross validation method. By fitting the total amino content to the spectra, the linear correlation coefficient was found to be 0.9860 and the linear correlation coefficient of cross validation was 0.9794. The develop method was used to determine amino acids in oyster and results show no obvious difference between using the developed NIR and the traditional amino acid analyzer, with a relative error of less than 1.0% and repeatability of 1.0% RSD(n= 10). This suggests that the new NIR method is feasible for high precision determining of amino acids in oyster.