Detection Of Dominant Spoilage Fungus In Apple Using Raman Spectroscopy And Imaging

Apple is rich in polyphenols,dietary fiber,organic acids and other components,with rich nutritional value,which is popular with people and as known of longevity fruit.However,apple is a kind of fresh food,which is easily infected by spoilage fungus in the process of storage and transportation,resulting in spoilage and deterioration,reducing the edible value of apple,causing huge economic losses and hindering the development of China's apple industry.Moreover,spoilage fungus can produce mycotoxins in the process of infection,causing toxin residues in apple and its products,which seriously endangers the health of consumers.Therefore,the detection of dominant spoilage fungus and their metabolites in apple is of great significance to ensure apple quality and reduce economic losses.Raman spectroscopy and imaging technology has a lot of advantages including non-destructive,high sensitivity and fast detection speed.It has been widely concerned and studied in the fields of medical diagnosis and treatment,environmental analysis and food detection,and has great application potential.Therefore,this study attempts to use Raman spectroscopy and imaging technology to detect the dominant spoilage fungus in apple.The specific research contents are as follows:(1)Characterization method research of the infection process of dominant spoilage fungus in apple by Raman chemical imaging.The infection of apple by spoilage fungus is a process of interaction between spoilage fungus and apple.The dynamic monitoring and in-situ analysis of the infection process of spoilage fungus provide new technical means and methods for revealing the interaction mechanism between spoilage fungus and fruits and vegetables,which is important to the early diagnosis of apple postharvest corruption and the prevention and control of apple biology.Firstly,the Raman spectra of cell wall and intercellular space of apple in different infection stages were collected,and the Raman spectra of standard samples were compared,and the Raman peaks were analyzed combined with relevant research literature;secondly,the average Raman spectra of apple in different infection stages were compared,and the changes of characteristic peaks intensity of cellulose,pectin and polysaccharide components during the infection process of spoilage fungus were analyzed;thirdly,the Raman spectra of apple cells were analyzed,and the main components in the cell wall and intercellular space were analyzed by Raman chemical imaging.The feature spectral peaks of cellulose,pectin and polysaccharide were selected to construct the pseudo color images of the distribution of these three components in the cell wall and intercellular space of apple,and the process of apple infected by spoilage fungus was observed from the micro perspective.Finally,principal component analysis(PCA)combined with linear discriminant analysis(LDA)was used to establish the discrimination models of different stages of apple tissue infection.Apple cells had response peaks at 526 cm^-1,875cm^-1,1080 cm^-1,1645 cm^-1 and 2946 cm^-1,which can be assigned to cellulose,pectin and lignin.The pseudo color images showed that these components were unevenly distributed in the cell wall and intercellular space of apple.With the aggravation of the infection degree of spoilage fungus,the cell structure was destroyed,cellulose,pectin and other components were decomposed,and the intensity of the characteristic Raman spectral peak showed a downward trend.PCA analysis showed that the Raman spectra of apples infected by spoilage fungus at different stages had clustering trend,and the recognition accuracy of the calibration set and prediction set were more than 90%.(2)Label-free detection of dominant spoilage fungus in apple using SERS technology.The infection of dominant spoilage fungus is an important reason for the loss of apple post harvest corruption.It is of great significance to detect and study the spoilage fungus quickly for the biological control and reduce the loss of apple corruption.Firstly,Gold nanorods(AuNRs)reinforced substrates were prepared using seed growth method;secondly,AuNRs were combined with negatively charged apple spoilage fungus by electrostatic adsorption and characterized;thirdly,SERS spectra of five dominant apple spoilage fungus were collected and Raman spectra were assigned;Finally,PCA,LDA,k-nearest neighbor(KNN)and support vector machine(SVM)were used to establish the identification model for dominant spoilage fungus in apple.The results showed that the SERS spectra of different spoilage fungus were different.The main Raman shifts of dominant spoilage fungus in apple were 496 cm^-1,687 cm^-1,1120 cm^-1,1200 cm^-1,1436 cm^-1,1479 cm^-1 and 1627 cm^-1.Anlysis found that chitin,protein,DNA,lipid and other cell components were the material basis of characteristic Raman spectra of spoilage fungus;three variable acquisition methods including interval partition,principal component extraction and feature spectrum extraction were selected to optimize the models.The recognition accuracy of the models established by various pattern recognition methods can reach more than 96%.The best modeling results were achieved using feature spectrum variables combined with BP-ANN and principal component variables combined with LDA,and the small batch experiment results showed that the initial recognition accuracy can reach 100%.(3)Label-free quantitative detection of apple patulin and alternariol using SERS.Patulin and AOH are the two most widely polluted toxins in apple and its products.It is important to establish a rapid and sensitive detection method for PAT and AOH in the products and to detect the contaminated products early.Firstly,AuNRs substrate with high stability and good reinforcing effect was synthesized;secondly,by optimizing the formation temperature and the volume of dropping liquid,the structure of coffee ring with regular shape,good enrichment effect and high enhancement effect was obtained;thirdly,the SERS spectra of mycotoxins in coffee ring region were collected,and the theoretical Raman spectra of PAT and AOH were calculated using density functional theory(DFT)and compared with the collected Raman spectra to determine the molecular vibration peaks of PAT and AOH;finally,the characteristic Raman spectra of PAT and AOH were selected by synergy interval(SI),genetic algorithm(GA)and uninformative variable elimination(UVE),and the quantitative detection model of PAT and AOH was established by partial least squares(PLS).It is found that the Raman spectra of the standard were basically consistent with the theoretical Raman spectra calculated by DFT.Due to the function of AuNRs,part of the spectral peak signals of PAT and AOH were enhanced,which improved the test sensitivity.Comparing the models established by different variable screening methods,the results showed that,for PAT detection,the best prediction result was obtained using the full spectra,the correction set correlation coefficient R_c was 0.9938,and the prediction accuracy R_p was 0.9831.For AOH detection,GA-PLS achieved the best modeling results,the calibration set correlation coefficient R_c was 0.9829,and the prediction set correlation coefficient R_p was 0.9808.In this study,the Raman spectroscopy and imaging detection methods of dominant spoilage fungus in apple were researched,and the interaction mechanism between spoilage fungus and apple tissue was explored from the cellular level.The rapid identification of five dominant spoilage fungus in apple and the rapid detection of their metabolites were realized.The research results provide a method reference for the research on the mechanism of spoilage fungus infection and the detection of fruits and vegetables,which is of great significance for ensuring the quality of fruits and vegetables and promoting the healthy development of fruit and vegetable industry.