Detection of Adulteration in Iranian Saffron Samples by 1H Nmr Spectroscopy and Multivariate Data Analysis Techniques Publisher



Dowlatabadi R^{1, 2} ; Farshidfar F^{3, 4} ; Zare Z² ; Pirali M^{2, 5} ; Rabiei M⁵ ; Khoshayand MR⁶ ; Vogel HJ^{1, 3} Show Affiliations
Source: Metabolomics Published:2017

Abstract

Introduction: The high market value of saffron (Crocus sativus L.) has made it an attractive candidate for adulteration. Safflower (Carthamus tinctorius L.) and tartrazine are among the most common herbal and synthetic foreign materials that may be added to pure saffron for the purpose of adulteration. In spite of encouraging advances achieved in the identification of adulteration in saffron samples, the lack of a simple method with sufficient power for discrimination of pure high grade saffron from meticulously adulterated saffron samples persuaded us to perform this study. Objectives: In this work, we show that 1H NMR spectroscopy together with chemometric multivariate data analysis methods can be used for the detection of adulteration in saffron. Methods: Authentic Iranian saffron samples (n = 20) and adulterated samples that were prepared by adding either different quantities of natural plant materials such as safflower, or synthetic dyes such as tartrazine or naphthol yellow to pure saffron (n = 22) composed the training set. This training set was used to build multivariate Principal Component Analysis (PCA) and Partial Least Squares Discriminant Analysis (PLS-DA) models. The predictive power of the PLS-DA model was validated by testing the model against an external dataset (n = 13). Results: PCA and PLS-DA models could both discriminate between the authentic and adulterated samples, and the external validation showed 100% sensitivity and specificity for predicting the authenticity of suspicious samples. Peaks specific to authentic and adulterated samples were also characterized. Proximity of samples with unknown adulteration status to the samples adulterated with known compounds in the PCA provided insight regarding the identity of the adulterant in the suspicious samples. Furthermore, the authentic samples could be distinguished based on their cultivation site. Conclusion: The present study demonstrates that the application of 1H NMR spectroscopy coupled with multivariate data analysis is a suitable approach for detection of adulteration in saffron specimens. Outstanding sensitivity and specificity of the PLS-DA model in discriminating the authentic from adulterated samples in external validation confirmed the high predictive power of the model. The advantage of the present method is its power for detecting a wide spectrum of adulterants, ranging from synthetic dyes to herbal materials, in a single assay. © 2017, Springer Science+Business Media New York.