摘要
直接取茶叶完整叶进行傅里叶变换近红外 (FT- NIR)光谱分析测定绿茶中氨基酸、咖啡碱和茶多酚的含量。茶叶的近红外光谱划分为 3个波段 ( :1 0 0 0 0— 71 4 3cm-1 :71 4 3— 5 376 cm-1 :5 376—4 0 0 0 cm-1 ) ,偏最小二乘法建立每种成分的数学模型 ,对不同波段的近红外分析结果进行讨论。在进一步优化波长范围的基础上获得了较好的数学模型 ,氨基酸、咖啡碱和茶多酚的校正值与化学值之间的相关系数分别达到 0 .994 ,0 .92和 0 .96 ,预测均方根误差分别为 0 .1 5 ,0 .31和 1 .6 7。
The free amino acids,caffeine and tea polyphenols in green tea were determined by Fourier transform near infrared (FT NIR) spectroscopy from the whole tea leaves.A partial least squares(PLS) algorithm was used to perform the calibration for each content in these 3 regions:Ⅰ(10000—7143cm -1 )and Ⅱ(7143—5376cm -1 )and Ⅲ(5376—4000cm -1 ),respectively.The best results were obtained in region Ⅲ.Better calibration models were built after further adjusting the wavelength regions.The correlation coefficients between the calibration and reference results are 0.994 for free amino acids,0.92 for caffeine and 0.96 for tea polyphenols.The root mean square error of prediction(RMSEP) are 0.15 for free acids, 0.31 for caffeine and 1.67 for tea polyphenols.
出处
《光谱实验室》
CAS
CSCD
2004年第5期940-943,共4页
Chinese Journal of Spectroscopy Laboratory
基金
浙江省自然科学基金项目 (编号 :2 0 2 0 81)
关键词
近红外光谱
偏最小二乘法
氨基酸
咖啡碱
茶多酚
Near Infrared Spectroscopy,Partial Least Squares,Free Amino Acids,Caffeine,Tea Polyphenols.