光谱学与光谱分析, 2019, 39 (11): 3519, 网络出版: 2019-12-02  

拉曼光谱结合偏最小二乘同时测定烟草中叶黄素和β-胡萝卜素含量

Simultaneous Detection of Lutein and β-Carotene in Tobacco by Using Raman Spectroscopy Combined with Partial Least Squares
作者单位
1 上海烟草集团有限责任公司技术中心, 上海 200082
2 华东理工大学化学与分子工程学院, 上海功能性材料化学重点实验室, 上海 200237
摘要
色素是烟草中的重要组成成分, 烟草的外观质量和内在品质都与色素相关。 色素主要分为绿色素、 黄色素和黑色素。 在生长期, 烟草中的绿色素主要是叶绿素, 成熟期的烟草中黄色素则主要为叶黄素和胡萝卜素, 黑色素则存在于成熟的烟叶中, 或在调制和发酵的过程中产生。 色素的检测对于烟草原料评价和成品质量都非常重要。 鉴于烟草色素基于液相色谱的传统检测方法耗时较长、 样品制备过程复杂, 以及拉曼光谱操作简单、 测定时间短、 能提供有关分子官能团信息的特点, 开发了一种应用拉曼光谱同时定量检测烟草中叶黄素和β-胡萝卜素的快检方法。 烟草样品的有机溶剂提取物, 密封在透明玻璃瓶中, 将激光聚焦于瓶内溶液样品直接测定拉曼光谱。 研究发现, 不同于常见的514和785 nm激发波长, 在短波455 nm激光作用下, 可获得更理想的拉曼信号, 光谱强度较高, 荧光干扰较低。 对色素萃取溶剂、 焦平面与光学平台间的距离等实验条件进行了优化。 考虑到不同检测日期仪器操作条件的变化可对拉曼光谱产生影响, 以溶剂峰为内标峰, 对光谱进行归一化处理, 以校正因测定条件不稳定而引起的物理干扰。 为了解决荧光干扰, 以及色素分子之间的拉曼光谱干扰问题, 采用偏最小二乘法(PLS)建立光谱分析的多元校正模型。 建模中对波长区域以及光谱预处理方法进行了优化。 研究结果表明, 光谱归一化处理显著降低了因物理因素产生的光谱干扰, 光谱求导运算对建模影响不大, 而波长选择能明显改善模型的预测能力。 利用798.2~1 752.8 cm-1波段建立叶黄素的PLS定量模型时, 预测效果最佳, 预测集的均方根误差(RMSEP)为6.68 μg·g-1; 对于β-胡萝卜素, 798.2~1 752.8 cm-1组合2 254.2~2 784.5 cm-1用来建模时得到的RMSEP最小, 为2.56 μg·g-1。 该方法操作简便、 耗时少、 结果准确可靠, 为烟草样本中色素的定量分析提供了一种切实有效的新途径。
Abstract
Pigments are important constituents in tobacco, that relate to appearance and quality of tobacco, meanwhile, the degradation products of pigments have great effects on the quality of aroma of tobacco. According to their color and property, pigments aregenerally divided into three categories: green pigment, yellow pigment and melanin. In growth, chlorophyll is the main green pigment in tobacco, while xanthophyll and carotene are the main yellow pigments in mature tobacco. Melanin consists in the progress of modulation and fermentation when tobacco leaves are ripe. Analysis of pigments in tobacco is important for evaluations of raw tobacco and the tobacco product quality. Traditional analytical methods of pigments in tobacco based on liquid chromatography, normally need long time and complex sample preparation process, while Raman spectroscopic method is simple to operate and takes short determination time, so that it could be able to provide information about molecular functional groups. Accordingly a method of simultaneous detection of lutein and β-carotene in tobacco was developed with Raman spectroscopy in the present work. A solution of organic solvent extraction from a tobacco sample was enclosed in a glass vial and was detected to collect its Raman spectrum by focusing the laser on the solution inside. The excitation light wavelength was optimized and 455 nm was selected, under which high Raman signals were obtained. Other experiment conditions, including extracting solvent of pigments and the distance between the focal plane and the optical platform were optimized. Because the spectra were measured in different dates and the measurement conditions could change, normalization was used to correctphysical interference due to the changes with a selected internal standard peak from the solvent spectrum. To solve problem of spectroscopic interference due to severe overlap between Raman signals of lutein and β-carotene, partial least squares (PLS) was utilized to build multivariate calibration models between Raman intensities and concentrations of pigments. The results showed that the normalization process could help to correct the changes of the measurement conditions, derivative calculations cannot improve the models but the models can be improved significantly by selection of wavelength regions. When spectral regionswere selected between 798.2 and 1 752.8 cm-1 for lutein, regions of 798.2~1 752.8 and 2 254.2~2 784.5 cm-1 for β-carotene, the built PLS models showed the best performances with the prediction errors of RMSEP being 6.68 and 2.56 μg·g-1, respectively. The results indicated that Raman spectra combined with PLS could supply a new way to the quantitative analysis of lutein and β-carotene in tobacco with advantages of easy operation, short time and reliable prediction.

葛炯, 孙琳, 沈晓洁, 沙云菲, 黄天雄, 杜一平, 张维冰, 谢雯燕, 姚鹤鸣. 拉曼光谱结合偏最小二乘同时测定烟草中叶黄素和β-胡萝卜素含量[J]. 光谱学与光谱分析, 2019, 39(11): 3519. GE Jiong, SUN Lin, SHEN Xiao-jie, SHA Yun-fei, HUANG Tian-xiong, DU Yi-ping, ZHANG Wei-bing, XIE Wen-yan, YAO He-ming. Simultaneous Detection of Lutein and β-Carotene in Tobacco by Using Raman Spectroscopy Combined with Partial Least Squares[J]. Spectroscopy and Spectral Analysis, 2019, 39(11): 3519.

关于本站 Cookie 的使用提示

中国光学期刊网使用基于 cookie 的技术来更好地为您提供各项服务,点击此处了解我们的隐私策略。 如您需继续使用本网站,请您授权我们使用本地 cookie 来保存部分信息。
全站搜索
您最值得信赖的光电行业旗舰网络服务平台!