ICP-OES analysis coupled with chemometrics for the characterization and the discrimination of high added value Italian Emmer samples

The elemental composition of 63 emmer samples produced in three different Italian territories, Monteleone di Spoleto (Umbria), Garfagnana (Tuscany) and National Park Gran Sasso-Laga (Abruzzo) were analyzed by Inductively Couples Plasma-Optical Emission Spectrometry (ICP-OES) combined with microwave-assisted digestion. The ICP-OES analytical method was validated to assess accuracy, precision and sensitivity. Recoveries were determined by the analysis of six genuine and six fortified emmer samples spiked at different concentrations. Recoveries (%) varied from 83 to 100% for the minor elements (Ba, Cu, Fe, Mn and Zn) detected in emmer in the 1−50 μg/g concentration range (acceptable recovery range 80–110%) and from 85 to 98% for the major elements (Ca, K, Mg and P) whose concentration in emmer was within 0.2−5 mg/g (acceptable recovery range 90–107%). The observed LOD and LOQ values ranged from 0.02 to 1.03 μg/gdry and from 0.07 to 3.44 μg/gdry, respectively, indicating a high sensitivity of the method. Analysis of Variance (ANOVA) was preliminarily applied to investigate the significance of the detected elements (Ba, Cu, Mn, Fe, Zn, Ca, K, Mg and P) as geographical markers. Eventually, Partial Least Squares-Discriminant Analysis (PLS-DA) was used to classify the samples according to their geographical origin. In order to validate the classification model, the 63 samples were divided into a training and a test set of 33 and 30 objects, respectively. The graphical investigation of the PLS-DA model highlighted that samples clearly group according to the geographical origin; this outcome is confirmed by the high accuracy achieved. In fact, the predictive model led to a correct classification rate of 100% for all the external samples. The investigation of the biplot and the application of the Variable Importance in Prediction (VIP) analysis indicated that all the elements except Cu contributed to the characterization of the three different geographical classes.