| چکیده انگلیسی مقاله |
Manipulation of lime juice composition undoubtedly has adverse effects on the quality of lime juice. Dilution of lime juice with water or a mixture of water and citric acid is a common adulteration type in the lime juice industry [1]. Although different methods such as chromatographic methods and biological methods have been previously used to detect lime juice fraud [2], finding a low-cost, quick, non-destructive and portable method will result in the quick screening of a large number of samples. Therefore, the current study aimed to evaluate the application of a portable visible-near infrared spectrometer (Vis-NIR; Link Square; 400-1000 nm) and chemometrics approach as a screening method for detection of lime juice adulteration. Twenty-four genuine industrial lime juice samples were used in the current study. These samples were authenticated based on the citric acid to iso-citric acid ratio as provided by the Association of the Industry of Juices and Nectars of the European :union: (AIJN). All the genuine samples were diluted with water at different levels (10, 15, 20, 25, 30, 35, and 40%) and subsequently, the Brix number was adjusted by the addition of citric acid. A handheld NIR device connected to a smartphone was used to analyze each sample (three replicates) in the reflectance mode. Following principal component analysis (PCA), the samples were not thoroughly clustered concerning their groups in score plots. However, the soft-independent modelling of class analogy (SIMCA) model as a class-modeling approach following smoothing and mean centering of data resulted in 100% sensitivity and 83% specificity in the prediction of genuine and adulterated samples, respectively. The overall accuracy of the model was calculated to be 91%. This level of accuracy provided empirical evidence of the potential of handheld NIR and chemometrics approach for the detection of lime juice adulteration, further investigations using classification models (e.g., partial least squares-discriminat analysis, PLS-DA) and regression models (e.g., partial least squares regression, PLSR) are required to confirm the promising results of the current study. |
