Combining with acid-base titration, HPLC, ATR-FTIR and chemometrics to study the effects of sulfur fumigation on medicinal and edible starchy samples

Abstract

Medicinal and edible starchy natural plants, commonly utilized as food, are often treated with sulfur fumigation (SF) to prevent spoilage during storage. However, SF can alter its active components, impacting overall quality. This study used acid-base titration and high-performance liquid chromatography (HPLC) to quantify residual sulfur dioxide (SO₂) and sulfite contents, while infrared (IR) spectroscopy was employed to characterize the SF starchy samples. The results indicated that SF changed the chemical components, with the degree of SF correlating positively with residual SO₂ and sulfite contents, and the main characteristic sulfite absorption peaks observed in the IR spectra at 2300–2000 cm⁻¹. Principal component analysis (PCA) of IR spectra revealed visible cluster trends among different degrees of SF starchy samples. Random forest, K-nearest neighbors, and linear discriminant analysis (LDA) were used to classify SF starchy samples. Combined with variable selection techniques of least absolute shrinkage and selection operator (Lasso) and SelectKBest, the optimized first derivative-LDA-Lasso model achieved a prediction accuracy of 98.88 %. This study demonstrated that the method has great potential to rapidly discriminate the SF starchy samples and other similar food products in the future.