VOCs determination by adsorbent-Raman system in food and botanicals

Volatile organic compounds (VOCs) emitted from foods and plants are typically analyzed by gas chromatography/mass spectrometry (GC/MS) or electronic-nose (E-nose) analysis. GC/MS is highly sensitive but lacks portability, while E-nose equipment is portable but has a narrow range of analysis. In this study, a Raman-based VOC detection system was developed along with multivariate analysis for spectral differentiation of VOCs. A 5 mu L droplet each of four representative VOCs (linalool, cis-3-hexenyl acetate, cis-3-hexen-1-ol, methyl salicylate) was pre-concentrated with Hayesep Q adsorbent in a sealed chamber. After a given collection time (20 and 60 min), Raman spectroscopy was used to measure the VOC eluates from each adsorbent. Two collection times were tested, and single VOCs versus VOC mixtures were compared. Finally, Raman spectral data were processed with baseline correction and normalization and analyzed with unsupervised and supervised learning techniques. After 20 minutes of collection, individual and mixed volatiles showed unique spectral features, and the multivariate techniques produced good discrimination among the VOCs. Raman spectroscopy was also used to measure mixtures containing three VOCs at five different concentrations, and when principle component regression was applied, a concentration-estimation curve with a high R-2 of 0.953 was generated. This system developed on VOC standards was tested on two types of tea (black and earl grey) to determine the differentiability of tea aromas, and several unique Raman spectra were observed from the tea samples. Therefore, the proposed Raman technique has proven to be effective at fast screening of multiple VOCs given off from foods and plants and is a good candidate for fast and portable applications.