Label-free surface enhanced Raman scattering spectroscopy for discrimination and detection of dominant apple spoilage fungus

Abstract

Fungal infection is one of the main causes of apple corruption. The main dominant spoilage fungi in causing apple spoilage are storage mainly include Penicillium Paecilomyces paecilomyces (P. paecilomyces), penicillium chrysanthemum (P. chrysogenum), expanded Penicillium expansum (P. expansum), Aspergillus niger (Asp. niger) and Alternaria. In this study, surface-enhanced Raman spectroscopy (SERS) based on gold nanorod (AuNRs) substrate method was developed to collect and examine the Raman fingerprints of dominant apple spoilage fungus spores. Standard normal variable (SNV) was used to pretreat the obtained spectra to improve signal-to-noise ratio. Principal component analysis (PCA) was applied to extract useful spectral information. Linear discriminant analysis (LDA) and non-linear pattern recognition methods including K nearest neighbor (KNN), Support vector machine (SVM) and back propagation artificial neural networks (BPANN) were used to identify fungal species. As the comparison of modeling results shown, the BPANN model established based on the characteristic spectra variables have achieved the satisfactory result with discrimination accuracy of 98.23%; while the PCA-LDA model built using principal component variables achieved the best distinguish result with discrimination accuracy of 98.31%. It was concluded that SERS has the potential to be an inexpensive, rapid and effective method to detect and identify fungal species.