摘要
应用傅立叶变换近红外漫反射光谱仪,采集340个样品光谱图,经优化挑选出具代表性的244个烟草样品建立了近红外光谱与烟草中的多酚含量间的数学模型,用建立的模型对20个样品进行预测,结果表明,各成分近红外预测值与实测值之间的平均标准偏差为0.10;且近红外预测值与化学法不存在显著性差异,可用近红外光谱快速检测烟草中的多酚含量,具有简便、快速、低成本、无污染以及样品的非破坏性等优点。
Correlation models for determination of polyphenols based on near-infrared (NIR) reflectance spectrum of 244 representative tobacco samples selected from 340 samples was established. The predicted results of 20 samples showed there were not markedly difference between predictive data and actual data. The mean deviation is 0. 10. The NIR analysis technique is simple,fast,low cost, without pollution, non-destructiveness sample, and can of tobacco. be used for quantitative analysis of polyphenols
出处
《光谱实验室》
CAS
CSCD
2008年第3期465-467,共3页
Chinese Journal of Spectroscopy Laboratory
基金
云南省烟草公司资助项目(05-17)
云南省"十五"科技攻关计划(2002NG01)
关键词
近红外光谱
烟草
多酚
NIR Spectroscopy
Tobacco
Polyphenols