UV-Visible Spectroscopy and Multivariate Classification as a Screening Tool to Identify Adulteration of Culinary Spices with Sudan I and Blends of Sudan I + IV Dyes

Abstract

This work propose a feasible, rapid, and simple method for detecting culinary spices adulterated either with Sudan I dye or blends of Sudan I + IV dyes at three concentration levels. The method is based on the use of UV-visible spectroscopy with multivariate analysis. Four types of spices were studied: three paprika varieties (mild, hot, and smoked) and a spice commonly consumed in Argentina called aji molido. Principal components analysis was firstly applied as an exploratory analysis and then, two classification techniques were used: K-nearest neighbors (KNN) and partial least squares-discriminant analysis (PLS-DA). Three classes were defined: unadulterated samples and adulterated samples with Sudan I or blends of Sudan I + IV dyes at 1, 2.5, and 5 ppm (mg L⁻¹). Classification techniques gave satisfactory results: between 89 and 100 % for PLS-DA and between 83 and 92 % for KNN. The sensitivity and specificity of the models were above 83 %. It has to be highlighted that none of the adulterated samples were assigned as unadulterated, which is very positive because of the implication that these results have on consumer health. The capability of detecting mixtures of Sudan dyes is a very important advantage because each Sudan dye generates different hazardous metabolites in human body so their toxicity may be enhanced by the simultaneous presence of such dyes.