Electronic nose combined with Chemometrics to assess emerging mycotoxin contamination in food

Electronic noses (e-noses) based on arrays of gas sensors are emerging as intelligent analytical devices simulating the human nose with high potential for rapid screening and classification of food samples according to their quality, safety and authenticity [1]. E-nose devices are composed of nonselective or semi-selective sensors that interact with volatile aroma compounds to produce electronic signals, which are processed using pattern recognition approaches and classification algorithms such as Principal Component Analysis (PCA), artificial neural networks, and other machine learning classifiers [2]. Mycotoxins, hazardous compounds produced by pathogenic fungi, pose a serious threat to consumer health. For this reason, preventing mycotoxin contamination is crucial for food safety. Regarding contamination of cereals by these naturally occurring toxins, fungal volatile metabolites can be used as an indicator of the presence of mycotoxins in cereals. In this context, the present study focuses on the development of an analytical approach to rapidly predict contamination from emerging mycotoxins, such as ergot alkaloids, in wheat using an e-nose based on metal oxide semiconductor sensors operating with an enrichment and desorption unit. A central composite design was applied to investigate the effects of extraction temperature and extraction time on the responses of selected volatile compounds; finally, the optimal conditions for the simultaneous extraction of the investigated volatiles were assessed by using the multicriteria method of the desirability functions. For data processing, the responses belonging to the most discriminant sensors were chosen, with the response signal being expressed as the conductance ratio G/G0 as a function of time for each sensor (where G and G0 are the conductance of a sensor in a detection gas and in clean air, respectively). The pattern recognition techniques used for data analysis were PCA, Hierarchical Cluster Analysis and Partial Least Squares-Discriminant Analysis. Results show that the electronic nose can successfully classify durum wheat samples discriminating between non-contaminated and contaminated samples below and above the EU regulatory level [3], demonstrating the potential of this intelligent sensory analysis device for high throughput screening of emerging mycotoxin contamination in wheat. Acknowledgements: METROFOOD-IT project has received funding from the European Union - NextGenerationEU, PNRR - Mission 4 “Education and Research” Component 2: from research to business, Investment 3.1: Fund for the realisation of an integrated system of research and innovation infrastructures - IR0000033 (D.M. Prot. n.120 del 21/06/2022) REFERENCES [1] Rabehi A et al., Appl. Sci., 14 (2024) 4506. [2] Tan J, Xu J, Artif. Intell. Agric, 4 (2020) 104. [3] Commission Regulation (EU) 2023/915 on maximum levels for certain contaminants in food and repealing Regulation (EC) No 1881/2006. OJEU L 2023, 119, 66, 103.