A chemometric approach to determine the phenolic compounds in different barley samples by two different stationary phases: A comparison between C18 and pentafluorophenyl core shell columns

• Two HPLC–DAD-MS methods have been developed with core shell C18 and PFP columns.
• C18 column method parameters have shown to be more suitable for the analyses.
• Phenolic compounds of waxy and non-waxy barley genotypes have been determined.
• A chemometric approach has been carried out to discriminate and classify the genotypes of barley samples.

Barley (Hordeum vulgare L.) is a cereal crop that has been cultivated since ancient times. However, its interest as nutritional food and as food ingredient is relatively new. Thus, in this study, the phenolic compounds of eighteen different varieties of barley (4 waxy and 14 non-waxy) grown under the same agronomic conditions in the same experimental field have been determined by HPLC–DAD-MS. Two new methodologies were developed using new generation superficially porous HPLC columns with different stationary phases: C18 and pentafluorophenyl (PFP). Twelve free phenolic compounds and eight bound phenolic compounds could be identified in barley samples in less than 22 min. The study of different method parameters showed that C18 column was more suitable for the analysis of phenolic compounds of barley. Hierarchical cluster analysis (HCA) was conducted in order to assess the different ability of the two different core shell HPLC columns in the discrimination between “waxy” and “non-waxy” varieties, and only HCA of C18 column could separate waxy and non-waxy genotypes.Significant differences in the content of phenolic compounds between waxy and non-waxy samples were found, being waxy barley samples the ones which presented higher content of free and bound phenolic compounds. Once the best discriminant HPLC column was established, principal component analysis (PCA) was applied and it was able to discriminate between “waxy” and “non-waxy” varieties; however it discriminated the barley samples based only in free phenolic compounds. Because of that, partial least squares discriminant analysis (PLS-DA) and Artificial Neural Networks (ANN) were carried out. PLS-DA and ANN permitted the classification of waxy and non-waxy genotypes from both free and bound phenolic compounds.