Analysis of Flavor Compound Variations in Coffee Beans Subjected to Different Roasting Methods by Using HS-SPME-GC-MS, E-nose, and HPLC

Abstract:

A comparative investigation of roasting methods by using catalytic infrared (CIR), electric infrared (EIR), and hot air (HA) was conducted to evaluate their distinct impacts on coffee bean flavor profiles through integrated analytical techniques, such as high-performance liquid chromatography, electronic nose, and headspace solid-phase microextraction gas chromatographymass spectrometry (HS-SPME-GC-MS). Significant inter-method variations (P<0.01) were identified in the contents of total sugars, polyphenols, lipids, proteins, chlorogenic acid, and trigonelline, whereas total acidity and caffeine levels remained statistically comparable (P>0.01). Orthogonal partial least squares discriminant analysis (OPLS-DA) of electronic nose data showed a prediction reliability of R2=0.858, whereas GC-MS characterization revealed 196 volatile compounds spanning 15 classes, with 104 shared volatiles establishing the most robust OPLS-DA model (R2=0.926). Six pivotal aroma discriminators-2,3-butanedione, 2,3-pentanedione, furfuryl acetate, 3,5-diethyl-2-methylpyrazine, methyl isovalerate, and hydroxyacetone-were quantitatively prioritized via odor activity value assessment, all demonstrating elevated concentrations in CIR-processed beans. Energy efficiency analysis further highlighted CIR’s superiority, exhibiting 73.87% and 23.68% cost reductions relative to HA and EIR, respectively. These findings collectively validate CIR as a dual-advantage roasting modality, simultaneously enhancing flavor complexity through targeted volatile compound preservation and optimizing energy utilization, thereby providing actionable insights for developing precision roasting technologies in specialty coffee manufacturing.