• Thienpont, Hugo
• Meulebroeck, Wendy
• Smeesters, Lien

Abstract

Food safety has become increasingly important in our society. Today a major concern of the agriculture industry is the screening of two specific types of food contaminants imposing a key threat to health, being acrylamide formed during high-temperature processing and mycotoxins produced by toxic fungi. This doctoral thesis investigates the use of spectroscopic sensing techniques to non-destructively detect these harmful products, avoiding their presence in the food chain. Particularly, our research contains three major parts, being the monitoring of the acrylamide formation in potatoes, the sensing of the presence of mycotoxins in cereals and the implementation of the spectroscopic detection techniques in real-time scanning configurations. We first investigate the use of spatially-resolved spectroscopy to monitor the acrylamide formation in potatoes. Our research reveals the influence of the acrylamide precursors on the light scattering behaviour, allowing to identify raw potatoes unsuited for high-temperature processing. In addition, we successfully validate this detection methodology in a proof-of-concept demonstrator, enabling an industrial integration in scanning-based sorting machines.To tackle our second objective, we investigate the use of reflection and fluorescence spectroscopy for the detection of respectively non-fluorescent and fluorescent mycotoxins. To sense the presence of non-fluorescent mycotoxins, we introduce a two-stage diffuse reflection measurement methodology. The proposed methodology allows to efficiently define the optimal detection wavelengths to sense the localized contamination-level, while enabling a pre-sorting of the samples. As a case-study, the sensing of deoxynivalenol in cereals is successfully demonstrated and integrated in a scanning-based proof-of-concept demonstrator. To detect fluorescent mycotoxins, we introduce two-photon induced fluorescence spectroscopy as a promising sensing tool and compare its performance to one-photon induced fluorescence spectroscopy. We develop a measurement configuration enabling the investigation of both types of fluorescence, define the optimal excitation wavelengths, and successfully demonstrate the detection of the localized aflatoxin-contamination.Finally, we propose an advanced optical scanning configuration, paving the way to integrate two-photon induced fluorescence in a commercial sorting machine. Particularly, we optimized the excitation laser power density and the fluorescence detection signal intensity by the integration of a tunable lens and a novel optical collection system. We conclude that our research demonstrates the use of advanced optical spectroscopy as a valuable tool to improve food safety, paving the way to an industrial, accurate and non-destructive detection.

Topics

• density
• impedance spectroscopy
• fluorescence spectroscopy
• light scattering