Use of polyphenolic fingerprints established by comprehensive two-dimensional liquid chromatography for the classification of honeys according to their floral origin

•Development of an LC×LC method for polyphenolic profiling of honey samples.•Two dimensional peak capacities around 2300 were achieved.•LC×LC optimized method gave 26 common polyphenolic peaks.•Samples were correctly classified according to floral origin with good resolution.•LC×LC is a promising technique to discriminate honey samples. In this work, the polyphenolic composition of honeys from three different floral origins (chestnut, heather, and thyme), coming from different geographical areas of Spain was investigated. First, samples were characterized in terms of total phenolic content (TPC) and antioxidant capacity, which was established by three different assays. The results revealed that the studied honeys presented similar TPCs and antioxidant capacities, with a wide variability within each floral origin. Next, a comprehensive two-dimensional liquid chromatography method was developed for the first time to establish polyphenol fingerprints of the three types of honeys, after optimizing the separation in terms of column combination and mobile phase gradient programs. After that, the detected common peaks were used for the construction of a linear discriminant analysis (LDA) model able to discriminate honeys according to their floral origin. The LDA model obtained was adequate for the classification of the floral origin of the honeys based on polyphenolic fingerprint data.