Coupled Extraction and Dynamic Headspace Techniques for the Characterization of Essential Oil and Aroma Fingerprint of Thymus Species

Natural essential oils and extracts were obtained from three different Thymus species (T. zygis subsp. gracilis (Boiss.) Morales, T. x citriodorus (Pers.) Schreb. and T. fontqueri (Jalas) Molero e Rovira) using hydrodistillation and supercritical fluid CO₂ extraction techniques, respectively. Furthermore, the aroma fingerprint of the plant materials was also characterized by collecting the volatiles through dynamic headspace-thermal desorption technique. In order to achieve that goal, the odor space was characterized by calculating odor values and by using an olfactory family model previously validated for both quantification and classification of volatile odorants. The obtained oils were characterized in terms of their organoleptic qualities as well as their chemical composition by gas chromatography coupled with mass spectrometry. A comparison amidst different species indicates that T. zygis and T. fontqueri show a phenolic type, mainly from high thymol and carvacrol content, respectively, while T. x citriodorus was richest in geraniol and citral. Furthermore, their organoleptic qualities also revealed odor similarities between T. zygis and T. fontqueri, both exhibiting a light herbaceous odor, in opposition to a strong lemon-herbaceous scent for T. x citriodorus. Hydrodistillation extraction yields ranged between 0.67% (w/w) for both T. zygis and T. fontqueri and 1.26% (w/w) for T. x citriodorus. In comparison, extraction yields for supercritical CO₂ samples were lower for T. x citriodorus and T. fontqueri (<0.05%, w/w), while for T. zygis it was obtained as 0.78% (w/w). Finally, CO₂ extracts were obtained for the first time for T. x citriodorus and T. fontqueri showing similar compositions to those obtained by hydrodistillation, although typically with lower phenolic and alcoholic content (especially for thymol, carvacrol, and geraniol). DHS analysis coupled with olfactory character models allowed us to classify both aroma and odor profiles of plants. In conclusion, coupling extraction techniques to obtain essential oils together with headspace volatile analysis appear to be a relevant approach for characterization of the fingerprint of plant materials, as well as establish a clear distinction amidst them.