Study On The Degradation Law Of Latent Aroma Substances In Tobacco Leaves During Natural Aging Based On Chemometric

In tobacco industry,aroma is an important indicator to evaluate the quality of tobacco.In the process of tobacco alcoholization,the precursor of aroma substances is a crucial factor affecting the quality of tobacco leaves.Despite extensive studies on aroma substances,the potential degradation mechanism of precursor and the formation mechanism of characteristic aroma substances are not completely elucidated.Therefore,a systematic study on the degradation rules of precursor in tobacco alcoholization has significant practical value.This paper combined chemometrics technology with plant metabolomics,and investigated the degradation rules of major aroma substances during tobacco alcoholization.Differences in degradation products of different aroma types(origin)during tobacco alcoholization were discovered,and a quantitative prediction model of important aroma substances was established by near-infrared spectroscopy.The research results can provide a scientific basis for quick determination of tobacco alcoholization quality,improvement of alcoholization process,and screening and development of high-quality tobacco flavor and fragrance.(1)In this study,the basic metabolic and synthetic processes of carotenoids,Amadori compounds,and sesquiterpenes in tobacco alcoholization were systematically summarized and investigated.Based on these studies,the sample preparation,detection methods,and analysis conditions of aroma substances were described.The simultaneous distillation extraction method was used to extract the aroma substances from tobacco leaves,and the GC-MS technique was used for detection.(2)To explore the impact of tobacco alcoholization on tobacco quality and aroma characteristics,a variance analysis was used to study the changes and differences in the degradation products of major aroma substances during tobacco alcoholization.The results showed that the degradation products of aroma substances from different origins and parts had different trends.Overall,the total content of carotenoid and Amadori compound degradation products showed an increasing trend,while the total content of sesquiterpene degradation products showed a decreasing trend in Yunnan Chuxiong,Guizhou Zunyi,and Liaoning Dandong,and an increasing trend in Hubei Enshi,Henan Xuchang,Fujian Sanming,and Shandong Rizhao.The overall trend in Hunan Chenzhou was not clear.(3)The chemical chemometrics analysis was used to analyze the different degradation products of alcoholized tobacco with different fragrances,and a variety of substances that contribute more to the fragrance of different fragrances were identified.It was found that substances such asβ-ocimene,β-caryophyllene,2-methyl-3-pentanone,β-phellandrene-2-ol,lycopene,β-damascenone,ionone,andβ-ionone were the major fragrance substances that contributed significantly to the alcoholization process of tobacco.In addition,it was found that there was a certain degree of similarity between different fragrances,which may be related to their similar content of certain substances.(4)The PLS regression model combined with near-infrared spectroscopy was used to construct quantitative analysis models for six aroma substances in alcoholized tobacco,including catechin,epicatechin,epicatechin gallate,scopoletin,isorhamnetin,and nicotine.This enabled the rapid prediction of the content of these six aroma substances in alcoholized tobacco.Analysis of the performance parameters of the aroma substance prediction model showed that the regression model fitted well,with the coefficient of determination of calibration(R~2cor)greater than 0.85 and the root mean square error of calibration(RMSEC)less than 1%.The root mean square error of prediction(RMSEP)was less than 1.2%,and the relative percent deviation(RPD)was greater than 2,except for epicatechin gallate.The prediction error of the model was within an acceptable range.The use of the PLS model to rapidly predict the content of aroma substances in tobacco at different alcoholization stages is helpful for controlling the degree and quality of tobacco alcoholization.