Innovative biosensors and analytical methods for the analysis and valorization of agri-food and agro-industrial waste

The presence of toxic contaminants is still a large analytical and environmental concern. This is especially true in areas with limited resources, like decentralized facilities. Reference laboratory analytical methods like chromatography and mass spectrometry provide exceptional sensitivity and selectivity for contaminant analysis, but the degree to which they rely on sophisticated instrumentation, trained personnel, and centralized infrastructure also limits their real-time and/or in situ applicability. The development and analytical validation of two sustainable, paper-based biosensing platforms for quantitative measurement of water quality are presented in this thesis. The first platform (elaborated in Chapter 2) is an integrated bioluminescent toxicity biosensor using entrapped Aliivibrio fischeri cells immobilised over a paper matrix. This system combines the use of smartphone imaging with an associated artificial intelligence application (Scentinel) and an on-board calibration model developed to compensate for variability due to different smartphone devices; this allows for model-independent quantitative analysis. The biosensor achieved a detection limit of 0.23 ppb for microcystin-LR and demonstrated high analytical sensitivity to environmentally relevant toxins. The second platform (elaborated in Chapter 3), NitriPad, is a ready-to-use origami colorimetric device based on the Griess diazotization reaction for nitrite (NO₂⁻) determination. The rationally engineered folding architecture allows for controlled mixing of reagents and initiation of reactions without the need for additional equipment or pipetting steps, thereby improving the reproducibility and robustness of the operational process. Analytical characterization of NitriPad yielded a limit of detection (LOD) of 0.27 mg L⁻¹ and a limit of quantification (LOQ) of 1.11 mg L⁻¹ when fortified drinking water samples with recoveries ranging from 87% to 110%. Furthermore, suitable linearity was established across environmentally relevant concentration ranges, and the platform was found to be amenable to smartphone imaging for quantification of results. White Analytical Chemistry (WAC) metrics were used to evaluate the sustainability of the two methods.