Chemometric intraregional discrimination of Chinese liquors based on multi-element determination by ICP-MS and ICP-OES

Abstract

The aim of this study was to illustrate the applicability of chemometric intraregional discrimination of three famous Chinese liquors in Sichuan province of China based on multi-element analytical results by atomic spectrometry. Twenty-one minor/trace and six macro elements were measured by inductively coupled plasma mass spectroscopy (ICP-MS) and inductively coupled plasma optical emission spectroscopy (ICP-OES), respectively. A two-step chemometric procedure was proposed, firstly finding out important element variables according to the random forests score of importance, and secondly predicting by random forests (RF) and support vector machine (SVM). It turned out that the liquor brand can be traced by constructing prediction models with 4 typical elements, Sr, Mg, Ca and Se. Meanwhile, an attempt at vintage prediction of two liquors from the same manufacturer was performed. These investigations showed that the brands of liquors in a small geographical area could be clearly identified by chemometric techniques combined with element analysis with only a few numbers of elements, and this will provide a convenient methodology for liquor differentiation and identification.