Analysis of wines elemental composition using ICP-MS/MS and artificial intelligence for classification from different origins

The elemental composition of wine can serve as a geographic fingerprint, contributing to both its market value and consumer perception. In this study, thirteen mineral elements (Al, V, Cr, Mn, Co, Ni, Cu, Zn, As, Mo, Cd, Ba, Pb) were subsequently quantified in 290 commercial wines from seven countries using Inductively Coupled Plasma Tandem Mass Spectrometry (ICP-MS/MS). Descriptive statistics showed asymmetric and non-normal distributions for all elements, with skewness values ranging from − 1.31 to 98.0, and p < 0.001 for most elements in the Kolmogorov-Smirnov test. Principal component analysis (PCA) and robust principal component analysis (ROBPCA) did not yield efficient sample classification based on origin, indicating the need for supervised methods. Supervised machine learning techniques such as support vector machine (SVM), random forest (RF), knearest neighbors (k-NN), and gradient boosting machine (GBM) were tested. Among them, RF achieved high classification performance for wine origin, with classification accuracy and kappa coefficient above 0.90. Barium was identified as the most relevant discriminant variable followed by cobalt, molybdenum and nickel. These findings highlight the potential of combining multi-element profiling with machine learning for reliable authentication and traceability of wines based on geographical origin. This could also be applied to beverages without geographical indication, as is the case for many wines from Brazil.