Delineating the extra-virgin olive oil aroma blueprint by multiple headspace solid phase microextraction and...

Comprehensive two-dimensional gas chromatography with parallel mass spectrometry and flame ionization detection (GC × GC-MS/FID) enables effective chromatographic fingerprinting of complex samples by comprehensively mapping untargeted and targeted components. Moreover, the complementary characteristics of MS and FID open the possibility of performing multi-target quantitative profiling with great accuracy. If this synergy is applied to the complex volatile fraction of food, sample preparation is crucial and requires appropriate methodologies capable of providing true quantitative results. In this study, untargeted/targeted (UT) fingerprinting of extra-virgin olive oil volatile fractions is combined with accurate quantitative profiling by multiple headspace solid phase microextraction (MHS-SPME). External calibration on fifteen pre-selected analytes and FID predicted relative response factors (RRFs) enable the accurate quantification of forty-two analytes in total, including key-aroma compounds, potent odorants, and olive oil geographical markers. Results confirm good performances of comprehensive UT fingerprinting in developing classification models for geographical origin discrimination, while quantification by MHS-SPME provides accurate results and guarantees data referability and results transferability over years. Moreover, by this approach the extent of internal standardization procedure inaccuracy, largely adopted in food volatiles profiling, is measured. Internal standardization yielded an average relative error of 208 % for the fifteen calibrated compounds, with an overestimation of + 538% for (E)-2-hexenal, the most abundant yet informative volatile of olive oil, and a -89% and -80% for (E)-2-octenal and (E)-2-nonenal respectively, analytes with a lower HS distribution constant. Compared to existing methods based on 1D-GC, the current procedure offers better separation power and chromatographic resolution that greatly improve method specificity and selectivity and results in lower LODs and LOQs, high calibration performances (i.e., R² and residual distribution), and wider linear range of responses. As an artificial intelligence smelling machine, the MHS-SPME-GC × GC-MS/FID method is here adopted to delineate extra-virgin olive oil aroma blueprints; an objective tool with great flexibility and reliability that can improve the quality and information power of each analytical run.