Classification of Dairy Products by Time-Correlated Single Photon Counting Fluorescence and Principal Component Analysis (PCA)

Abstract

A transient fluorescence spectroscopy system based on time-correlated single photon counting (TCSPC) technology was constructed. Time-resolved fluorescence spectra of pure, modified and fresh milk were obtained using this system. A method for differentiating these dairy products was established by principal component analysis (PCA). The measured time-resolved fluorescence spectra of the dairy products were directly substituted into the principal component analysis model without fitting. The results showed that the two largest principal components, PC1 and PC2, had variance contribution values of 48.18% and 32.1%, respectively. The results were plotted as a two-dimensional graph of PC1 × PC2. The results showed that the coordinates of the dairy products were clearly distinguished in the two-dimensional plot, with pure milk in the fourth quadrant, modified milk in the second and third quadrants, and fresh milk in the first quadrant. It is inferred that the time-resolved fluorescence spectra were changed due to the different heat treatment of the dairy products. Compared with fluorescence spectroscopy, time-resolved fluorescence spectroscopy is less susceptible to instrumental variation. Data from different instruments can be compared. The developed method does not require any processing and fitting of fluorescence lifetimes to milk, which has the advantages of easy identification, simple operation, and short identification time.