Fingerprinting cromatografico ad alto potenziale informativo mediante gas cromatografia bidimensionale comprehensive accoppiata alla spettrometria di massa a tempo di volo: caratterizzazione chimica di off-flavours in nocciole di differenti origini geografiche

Hazelnuts are a crop of great economic relevance; their consumption is driven by sensory pleasure being one of the most valuable ingredients of many comfort food (chocolate bars and chocolate spreads). In this study, hazelnuts sensory defects deriving from characteristic patterns of odorants connoted by unpleasant descriptors (fatty, rancid, mould, stale etc..) have been exploited to derive analytical strategies for effective and objective classification in the perspective of quality control and products acceptance. Head Space Solid Phase Microextraction (HS-SPME) sampling coupled with Gas Chromatography- Mass Spectrometry were applied to identify within the highly complex volatile metabolome of raw hazelnuts informative patterns of chemicals to be adopted as classification probe(s). A strategy capable of discriminating raw hazelnuts with off-flavours notes was developed and validated; 7 odorants belonging to different chemical classes are capable of correctly classify bad samples from those with acceptable quality. They are: hexanal, heptanal, octanal, acetic acid, 1-pentanol, 1-hexanol and 1-heptanol. Thanks to the effective separation and detection power of comprehensive two-dimensional gas chromatography (GC×GC) equipped with a loop-type thermal modulator and coupled with Time of Flight Mass Spectrometry (TOF MS) a more complete description of volatiles signatures of raw hazelnuts with codified sensory defects (rancid, rancid-solvent, rancid-stale, mould and mould-rancid-solvent) was obtained. About 120 targeted compounds were identified and, on the basis of their characteristic distribution, an improved discrimination between OK-KO hazelnuts was achieved. To combine the effective classification potential and the informative power emerged in the previous phases, the principles of method translation were exploited to transfer the volatile metabolome fingerprinting method from the loop-type thermal modulated to a reverse-inject differential flow modulated (FM) GC2GC-MS/FID platform. The information related to volatiles patterns achieved with thermal GCGC-MS system was almost preserved despite of the lower sensitivity showed and reduced separation performances. A 68% of the analytes involved in the OK-KO classification were successfully detected while 77% and 85% were mapped for respectively mould and mould-rancid-solvent classes discrimination.