Quantitative volatilomics by GC×GC–MS/FID parallel detector signal fusion: tracking potent odorants and quality markers in hazelnuts

Volatile metabolite profiling of complex food matrices requires analytical strategies capable of combining confident structural identification with accurate quantification. Comprehensive two-dimensional gas chromatography (GC×GC) coupled with parallel flame ionization detection (FID) and mass spectrometry (MS) represents a powerful platform: MS provides structural confirmation, while FID ensures robust and reliable quantification. In this study, we address the challenge of fusing MS and FID chromatographic data to enhance pattern recognition during template matching, while enabling the simultaneous quantification of a large set of quality-informative features. MS-derived spectral similarity guides feature matching across chromatograms, reducing mismatches and improving both specificity and selectivity in the extraction of FID responses for accurate quantification. This approach was applied to hazelnuts, a high-value ingredient in the confectionery industry, whose volatile metabolome is influenced by cultivar, geographic origin, post-harvest treatments, microbial contamination, and oxidative degradation1. The ability to reliably monitor informative markers, including key odorants, within a single analytical run is of considerable industrial relevance. Hazelnut samples collected over four production years exhibited variability in MS response and chromatographic misalignments, leading to inconsistencies in two-dimensional peak patterns2. The application of MS-driven data fusion minimizes feature mismatches, reducing false negatives among the 441 detectable compounds in raw hazelnuts compared to FID-only processing, while also decreasing false positives and thus improving method specificity and selectivity. Data fusion additionally reduces processing time and facilitates metadata transfer. Following pattern recognition, FID signals are extracted for quantification using external calibration or predicted FID response factors. Quantitative volatilomics based on parallel detector signal fusion enables tracking of odorant changes across harvests and throughout shelf life, while supporting robust marker discovery for industrial quality assessment.