Gas chromatography-mass spectrometry coupled to PARAFAC-SIMCA for the authentication of acacia honey

Abstract

Honey's botanical origin heavily influences its composition and market value, driving the need for reliable authentication methods. Monofloral honeys, derived from a single plant species, have distinct flavor and nutraceutical properties, increasing demand and making them susceptible to counterfeiting. Pollen analysis is the standard for botanical origin authentication but is labor-intensive and requires extensive regional pollen data. This study introduces a chemometric approach to discriminate honey types using volatile profiles obtained through HS-SPME-GC-MS. A dataset of 98 samples, including 49 acacia honeys, was analyzed. Second order GC-MS data were deconvoluted into temporal, spectral, and concentration profiles via Parallel Factor Analysis (PARAFAC). The resulting concentration profiles were used to develop a SIMCA (Soft Independent Modelling of Class Analogy) model, using acacia honey as the target class. The proposed strategy allowed the discrimination of acacia honey from samples of diverse botanical origin, achieving high sensitivity (86%) and specificity (92%), with an overall accuracy of 91%. Further analysis of the data revealed that acacia honeys exhibited higher levels of linalool oxide, furfural, benzaldehyde, and hotrienol, and lower levels of lilac aldehydes (B, C, D), and 1-p-menthen-9-al, compared to other types. The developed method permits honey authentication and compositional analysis, contributing to understanding honey's chemical diversity and ensuring traceability.