Aromatic Fingerprints: VOC Analysis with E-Nose and GC-MS for Rapid Detection of Adulteration in Sesame Oil

Abstract

Food quality assurance is an important field that directly affects public health. The organoleptic aroma of food is of crucial significance to evaluate and confirm food quality and origin. The volatile organic compound (VOC) emissions (detectable aroma) from foods are unique and provide a basis to predict and evaluate food quality. Soybean and corn oils were added to sesame oil (to simulate adulteration) at four different mixture percentages (25–100%) and then chemically analyzed using an experimental 9-sensor metal oxide semiconducting (MOS) electronic nose (e-nose) and gas chromatography–mass spectroscopy (GC-MS) for comparisons in detecting unadulterated sesame oil controls. GC-MS analysis revealed eleven major VOC components identified within 82–91% of oil samples. Principle component analysis (PCA) and linear detection analysis (LDA) were employed to visualize different levels of adulteration detected by the e-nose. Artificial neural networks (ANNs) and support vector machines (SVMs) were also used for statistical modeling. The sensitivity and specificity obtained for SVM were 0.987 and 0.977, respectively, while these values for the ANN method were 0.949 and 0.953, respectively. E-nose-based technology is a quick and effective method for the detection of sesame oil adulteration due to its simplicity (ease of application), rapid analysis, and accuracy. GC-MS data provided corroborative chemical evidence to show differences in volatile emissions from virgin and adulterated sesame oil samples and the precise VOCs explaining differences in e-nose signature patterns derived from each sample type.

Design of IoT-based Volatile Organic Compounds Monitoring System

“…Commonly used sensor modules rely on fixed heating resistors and fixed voltage for sensor heating. Additionally, these devices are often constrained by microprocessors and circuitry, and the presence of multiple functionalities tends to reduce the range of detectable gas types [4][5][6][7] .…”

Complementary assessment of nano‐packaged garlic properties by electronic nose

“…In contrast, E‐Nose is mapped out with several non‐selective sensors that link to odor molecules (Xing et al., 2023 ). The result of such connections is a diverse signal's classes which are sent to a computer to recognize their patterns by multivariate statistics (Aghili et al., 2023 ). A broad list of E‐Nose applications can be found in the agricultural industry (Seesaard et al., 2022 ), environmental monitoring (Kumar et al., 2023 ), food analysis (Oates et al., 2022 ), and quality control (Pulluri & Kumar, 2022 ).…”

Identification of Milk Adulteration in Camel Milk Using FT-Mid-Infrared Spectroscopy and Machine Learning Models

“…Ana et al employed a convolutional neural networkbased image method to identify water adulteration in milk, achieving an accuracy of 93% [20]. Nadia et al successfully established a sesame oil adulteration model using an electronic nose combined with SVM and ANN, where the SVM model exhibited slightly superior sensitivity and specificity compared to the ANN model [21]. These relatively new technologies offer advantages such as excellent generalization capabilities, the ability to model non-linear data, high scalability, and ease of training [22].…”