A comparative study of SPME GC-MS vs SPME-MS approaches for characterization of honey source

Abstract

Honey is greatly appreciated by consumers not only for its nutritive properties and sweet taste, but also for its pleasant aroma. Different unifloral, multifloral and honeydew honeys are marketed nowadays, and their characterization for authentication purposes or to increase their added value represents a challenge for the analyst. Honey aroma is dependent on its volatile composition, which in turn is influenced by the nectar or honeydew collected by bees as well as by beekeeping practices, processing and storage conditions, etc. The high resolution and sensitivity provided by Gas Chromatography coupled to Mass Spectrometry (GC-MS) make of this technique the tool of choice for the analysis of the complex mixtures of volatiles of different functionality, and usually present at very low concentrations, in honey. However, GC-MS analysis of honey volatiles requires a previous step in which these compounds are extracted from non-volatile honey matrix. Solid-Phase Microextraction (SPME) has been shown to be a simple, rapid and affordable technique for the extraction and preconcentration of honey volatiles (1). Furthermore, the selectivity of SPME extraction can be modulated by the fiber coating selected (2). On the other hand, the development of non-separative approaches based on mass spectral fingerprints is recently gaining wide attention for the rapid and unbiased classification of food samples (3). However, up to date, they have scarcely been applied for the characterization of honey source (4). The objective of this work was to evaluate and compare the performance of two approaches for the characterization of honey source: (i) in the ‘conventional approach’, multivariate statistical analysis was applied to semiquantitative data gathered by SPME GC-MS. A total of 40 commercial honeys of 13 different sources were analyzed (1) using two different SPME fiber coatings (polyacrylate and carboxen/polydimethylsiloxane). (ii) In the ‘mass spectral fingerprint approach’, after corresponding data pretreatment (removal of interferences and normalization), chemometrics was applied to combined MS spectra of volatiles sampled by SPME. Classification of honeydew honeys and of several unifloral honeys (citrus, eucalyptus, rosemary and acacia) was possible by both approaches. However, mass to charge ratios and volatiles pointed out as markers of honey source in the first and second approaches, respectively, were only sometimes related. Similar conclusions were obtained irrespective of the SPME fiber coating considered. To conclude, the usefulness of both SPME GC-MS and SPME-MS approaches for the classification of honey source has been proved. References: (1) Soria AC, Martínez-Castro I, Sanz J. (2003). J Sep Sci 26, 793–801. (2) Soria AC, Sanz J, Martínez-Castro I. (2009). Eur Food Res Technol 228, 579–590. (3) Pérès C, Begnaud F, Eveleigh L, Berdagué J-L. (2003). Trends Anal Chem 22, 858-866. (4) Aliferis KA, Tarantilis PA, Harizanis PC, Alissandrakis E. (2010). Food Chem 121, 856-862.