Gas nanosensors for quality assessment of food products

This PhD project arises from the growing awareness of food quality and safety on the part of citizens and institutions. The increase in the population to feed, the lengthening of the production and distribution chains and the socio-economic risks of a poor diet make it crucial to monitor the quality of food from the producer to the consumer. Traditional methods (sensory panels and laboratory analytical techniques) are too expensive and above all slow to evaluate the quality of fresh foods that deteriorate over the course of a few hours. In this context, it is crucial to develop monitoring devices that are cheap, rapid and non-invasive, in order to be able to evaluate the quality of food products extensively and constantly. Solid-state gas sensors are an ideal candidate, as they are inherently non-invasive and inexpensive. In this context, the project focused on chemoresistive sensors based on semiconductor metal oxides, which are among the simplest and most performing, and have the advantage of being sensitive to almost all gases and VOCs. Initially, nanostructures of different materials (n- and p-type semiconductors) and of different morphologies (nanowires and nanosheets) were studied in order to investigate the performance of individual sensors. In this way, some devices have been optimized with respect to the detection of possible biomarkers of the degradation of specific foods. The sensors have demonstrated a rapid response (from a few seconds to a minute), an intense response and above all a very low detection limit (less than 1ppmv, in some cases a few tens of ppbv), important for agri-food applications. This approach is the simplest since it requires a single sensor that is selective towards a certain molecule (ammonia, ethylene...) which can be considered the only important information parameter in a certain application. In most cases, however, the gaseous emission of a food is composed of a large quantity of volatile compounds, and the low selectivity of resistive sensors makes it difficult to discriminate the molecules most informative regarding the degradation process. For this reason, in the second part of the PhD we used the sensors developed up to then to create electronic noses. Exploiting the dimensions of nanostructures, we have developed a new concept of thermal electronic nose, i.e. with sensors of the same material, but within a thermal gradient. In this way, by exploiting multivariate statistical analysis and machine learning techniques, the devices acquired a greater ability to discriminate and quantify the different gases. The electronic noses have shown that they can perfectly recognize the different gases tested (100%) and estimate their concentration with an error of a few ppmv. Measurements in the laboratory are very useful for testing the performance parameters of sensors and electronic noses, as they make it possible to evaluate the correctness of the classification and the error in estimating the concentration of any gas. On the other hand, measuring the emissions of fresh food is different, as the concentrations are not known, and therefore a different approach is needed. The final stage of the project involved using electronic noses to assess the freshness of certain agri-food products. As the developed sensors were particularly sensitive to ammonia, it was decided to study the degradation of meat and fish, where this gas is an important marker. The electronic noses have been able to accurately recognize the meat from the fish (> 95%), and evaluate the state of degradation by giving a very accurate estimate of the microbial count (>95%), responding in a very short time (tens of seconds). The miniaturized electronic noses developed during this PhD project have therefore successfully demonstrated to be a rapid and non-invasive cross-sectional tool for assessing the freshness of agri-food products.