Rapid Non-destructive Inspection Of Hazard Fungal Contamination In Peanuts

Peanut products were susceptible to the change of temperature and humidity during storage.The peanuts were easily infected by hazard fungal species and produce the most potent mycotoxin,then endangering human health.Current methods for fungi contamination determination in peanuts are usually labor-intensive and time-consuming.The quality of peanut products were rapidly and non-destructively assessed for storage and edibility safety.This paper used characteristic volatiles and spectra information arising in peanuts during storage as the pointcut.Firstly,clean and fresh peanuts were sterilized by Co-60 and inoculated individually with five common hazard fungal species in grains,namely A.flavus 3.17,A.flavus 3.3950,A.parastiticus 3.3950,A.parastiticus 3.0124,and A.ochraceus 3.6486.The samples were then incubated for 9 days under suitable conditions(26 ℃,RH 80%).Secondly,the electronic nose was used for the collection of volatile odor information from peanut samples stored for 0,3,6 and 9 d,respectively.Spectra information were collected from peanut samples in the wavenumber range 12000 cm-1 to 4000 cm-1 and 4000 cm-1 to 600 cm-1at different time by fourier transform near infrared spectroscopy(FT-NIR)and fourier transform middle infrared spectroscopy(FT-MIR)during the inoculation.Finally,qualitative and quantitative models for the determination of harmful fungi contamination in peanuts were established by multivariate statistical analysis method.The main conclusion of this work were obtained as follow:1,The principal component analysis(PCA)results showed that electronic nose could effectively distinguish different fungal infection levels of peanut samples.The discrimination rate of linear discriminant analysis(LDA),partial least squares discriminant analysis(PLS-DA)and support vector machines discriminant analysis(SVM-DA)to peanut samples under different storage periods have been reached 100%.SVM models were developed to predict the number of colonies of peanut samples with the coefficient of determination of the validation set(RP2)of 0.8547,root mean square error of cross-validation(RMSECV)of 0.2631 Log CFU/g and residual predictive deviation(RPD)of 2.31.To distinguish different fungal species of A.flavus 3.17,A.flavus 3.3950,A.parastiticus 3.3950 and A.parastiticus 3.0124 by using LDA,PLS-DA and SVMDA analysis models,the discriminant rate have been reached 87.5%.The peanut samples infected with fungal species might produce more aromatic compound,hydrocarbon compounds,amino compounds and alkanes by loading analysis.The results demonstrated that the electronic nose technology could be rapid determination of fungal contamination levels in peanuts,which could realize quality and safety control during storage of peanuts.2,The rapid analysis models of fungal contamination in peanut samples were established based on FT-NIR technology.Through analyzing the raw averaged spectrum of peanut samples infected by 5 different fungal species,the chemical bonds of protein,fat and carbohydrate in peanut samples appeared various degrees of fluctuation.The PCA analysis showed that the inoculated different fungal species of peanuts can be effectively distinguished during different storage periods.For the classification of peanut samples with different storage periods,the correct rate of 97.5%was obtained by LDA models.PLSR models were developed to predict the number of colonies of peanut samples with Rp2,RMSECV and RPD were 0.8741,0.2760 Log CFU/g and 1.92,respectively.To distinguish 4 fungal species by using LDA models,the discriminant rate was 87.5%.The results demonstrated that the FT-NIR technology could be used as a reliable analytical method for rapid determination of fungal contamination levels in peanuts.3,The loading analysis of peanut samples were established based on FT-MIR technology.Through analyzing the peak spectrum of samples appeared various fluctuation during different storage periods.For the classification of peanut samples with different storage periods,the correct rate of 87.5%was obtained by PLS-DA models.PLSR models were developed to predict the number of colonies of peanut samples with Rp2,RMSECV and RPD were 0.7803,0.3580 Log CFU/g and 1.76,respectively.Qualitative and quantitative analysis models for the determination of harmful fungi contamination in peanuts by using electronic nose,FT-NIR and FT-MIR technology during storage.The results showed that these 3 technologies could be used as reliable analytical methods for rapid determination of fungal contamination in peanuts,in which the overall results of electronic nose were better,and FT-MIR technology were relatively poor.The reasons include inhomogeneous fungal species and mycotoxin distribution in peanut samples,and so on.In order to promoting the performance of model,strict limitation of consistency and representation of samples,as well as feature extraction methods should be implemented for further research.