Simultaneous Detection of Glucose and Xylose in Tobacco by Using Partial Least Squares Assisted UV-Vis Spectroscopy
Li Yu,Zhang Ke-can,Peng Li-juan,Zhu Zheng-liang,He Liang
DOI: https://doi.org/10.3964/j.issn.1000-0593(2024)01-0103-08
2024-01-01
Spectroscopy and spectral analysis
Abstract:Tobacco contains a variety of sugars, including glucose, fructose, maltose, sucrose, xylose, etc. These sugars account for 25% to 50% of the dry weight of tobacco and are an important component and a crucial criterion for measuring the intrinsic quality of tobacco. Different sugars produce different products during the combustion process of cigarettes, which directly or indirectly affect the aroma and taste of cigarettes. Therefore, establishing methods for determining various monosaccharides in tobacco is relevant for the quality control of tobacco products. Traditional detection methods such as Ferring's reagent method and continuous flow analysis can only determine the content of water-soluble sugars or total sugars in tobacco, and cannot quantify single sugar components. However, gas chromatography and liquid chromatography can detect different monosaccharides, they have limitations such as troublesome sample pre-treatment process and long detection time. In this work, based on the traditional phloroglucinol chromogenic method for determination of xylose, the method was improved based on the color development mechanism to solve the technical problem of poor color development of glucose under this detection system, and the precise determination of glucose and xylose content in tobacco and cigarette products can be achieved and efficiently by UV-visible spectroscopy. When the above two sugar fractions are measured simultaneously, the UV spectra can affect each other's measurement accuracy due to their similar peak positions. A multivariate correction model for spectral analysis was developed using partial least squares regression (PLS), aided by a modified phloroglucinol chromogenic method to simultaneously determine glucose and xylose contents in tobacco and its products. The results showed that the linearity of the modified phloroglucinol method was satisfactory for glucose and xylose solution at concentrations in the range of 0. 05 similar to 0. 4 and 0. 10 similar to 0. 80 mmol . L-1, respectively, and the limits of detection and limits of quantification of the method were 0. 001 7, 0. 005 7 mmol . L-1 and 0. 007 2, 0. 024 mmol . L-1, respectively. The intra-batch coefficients of variation were in the range of 0. 69%similar to 3. 03%, and the spiked recoveries were in the range of 96. 72%similar to 102. 85%. The lignin content in the extracts did not interfere significantly with the determination results. For the mixed sugar solution composed of glucose and xylose, the external test was used to evaluate the model effect, and the correlation coefficient R-2 = 0. 994 7 between the predicted and theoretical values of the external test set was within 10% relative deviation. It can be seen that this method can quickly and accurately determine the content of glucose and xylose in tobacco and tobacco products and provide an effective method for monosaccharide analysis in the tobacco industry.