近红外光谱法快速测定河南烟叶中木质素的含量

[目的]实现对河南烟叶中木质素含量的快速测定。[方法]首先改进了烟叶中木质素含量的测定方法,并对该方法测定的准确性进行了评定(对同一样品,8次平行测定的RSD值为3.58%),随后用此方法测定了河南地区95个初烤烟叶样品中的木质素含量,然后用近红外光谱仪采集相应的光谱,进行二阶微分处理和平滑预处理后,用偏最小二乘法和完全交互验证方式建立相应预测模型。[结果]试验建立的烟叶中木质素预测模型的相关系数达到96.13%,内部交叉检验均方根误差(RMSECV)为0.104,说明烟叶中木质素含量和近红外光谱之间存在较好的相关性。[结论]采用近红外光谱法可以快速预测河南烟叶中木质素含量。

Rapid Determination of Lignin Content in Henan Tobacco Leaves by Near Infrared Spectroscopy

Objective] To realize the repaid prediction of lignin content in tobacco leaves of Henan Province .[ Method] Determination meth-od of lignin content in tobacco leaves was improved and the accuracy of the method was evaluated ( for the same sample , the RSD value of 8 parallel tests was 3.58%).The lignin content in 95 samples of early baked tobacco leaves in Henan Province was determined by this method. Then, corresponding spectra were collected by the near infrared spectrometer.After the second order differential treatment and smooth pre-treatment, corresponding prediction model were established by partial least square method and fully interactive authentication method.[Re-sult] The correlation coefficient of lignin prediction model was 96.13%, RMSECV was 0.104.[ Conclusion] There is good correlation be-tween lignin content and near infrared spectrum, and the lignin content in Henan tobacco can be predicted by near infrared spectroscopy.