Reza Jahani - Department of Toxicology and Pharmacology, School of Pharmacy, Shahid Beheshti University of Medical Sciences, Tehran, IR Iran- Wageningen Food Safety Research, Wageningen University and Research, Akkermaalsbos ۲, ۶۷۰۸ WB, Wageningen, The Netherlands- Foo
Yannick Weesepoel - Wageningen Food Safety Research, Wageningen University and Research, Akkermaalsbos ۲, ۶۷۰۸ WB, Wageningen, The Netherlands
Martin Alewijn - Wageningen Food Safety Research, Wageningen University and Research, Akkermaalsbos ۲, ۶۷۰۸ WB, Wageningen, The Netherlands
Hassan Yazdanpanah - Department of Toxicology and Pharmacology, School of Pharmacy, Shahid Beheshti University of Medical Sciences, Tehran, IR Iran- Food Safety Research Center, Shahid Beheshti University of Medical Sciences, Tehran, IR Iran
Farzad Kobarfard - Department of Pharmaceutics, School of Pharmacy, Shahid Beheshti University of Medical Sciences, Tehran, IR Iran
Mehrdad Faizi - Department of Toxicology and Pharmacology, School of Pharmacy, Shahid Beheshti University of Medical Sciences, Tehran, IR Iran

Introduction: Since the incidence of food adulteration is on the rise, finding a rapid, accurate, precise, low-cost, user-friendly, high-throughput, ruggedized and ideally portable method is a valuable issue in food sciences. Near-infrared spectroscopy (NIRS) in combination with chemometrics approach has been suggested to have potential for continuous monitoring of the market. This study will focus on the feasibility of a benchtop and a portable NIRS in the detection of citric acid-adulterated lime juice samples.Methods: Collected lime juice samples were divided into two groups of genuine and adulterated juices based on citric acid and iso-citric acid content. After acquiring NIR spectra (1000–2500 nm in FT-NIRS and 740–1070 nm in portable NIRS), obtained data were subjected to different algorithms in order to classify genuine and adulterated juices.Results: Principal component analysis of spectral data from FT-NIRS did not generate two separate clusters, but a noticeable distinction between genuine and adulterated samples was obtained along second PC in the portable NIRS data. Considering the result of partial least squares discriminant analysis on benchtop NIRS data, smoothing followed by log 1/T and standard normal variate transformation of the data on the training set delivered satisfactory results on the prediction of test set (100% sensitivity and 94% accuracy). Same accuracy and 91% sensitivity were delivered by multiplicative scatter correction in combination with 2nd derivative transformation for portable NIRS data.Conclusion: The promising results allowed us to conclude that NIRS technology coupled with the appropriate multivariate classification models enables fast detection of citric acid-adulterated lime juices. Since limited samples were used in this study, further model validation using lime juice from a wider range of variability is required to confirm the promising results of this paper.

Lime, Portable NIR, FT-NIR, Adulteration, Chemometrics, PLS-DA