Abstract
Introduction
Nowadays, air pollution still remains a critical issue. Global agreements have been signed by several countries to limit emissions of air pollutants [1], introducing increasingly stringent legislation. Moreover, national and supranational organizations, such as the UN and the European Union, publish updated reports on the situation on an annual basis. In recent years, research has focused not only on the use of renewable energy sources for sustainable development, but also on increasingly simple and reliable methods for detecting and mapping the concentration of toxic and polluting gases [2]. To achieve this goal, it is essential to develop sensors capable of easily acquiring and providing data in real time.
In this work we present the results obtained in the project CHEARIA, a collaborative activity between Bruno Kessler Foundation (FBK) of Trento, the local environmental protection agency (APPA) and several high schools of the Trentino-Alto Adige Region, as part of the national training program "work-based learning". The program was mainly focused on the development of chemoresistive gas sensor arrays produced at FBK, and their use to monitor the city air quality. Data were collected by exploiting the Internet of Things (IoT). The obtained results were compared with those obtained with the certified systems of the APPA agency.
Material and Method
Different MOX nanopowder were prepared and tested: SnO2, ZnO nanograin, ZnO nanorods, WO3, TiO2 and solid solution of them. The nanostructured MOX were synthesized through sol-gel method in hydroalcoholic solutions and then calcined. The chemical composition, morphology and crystalline structure of the synthesized nanopowders were investigated.
The sensing materials were then screen printed onto silicon microheaters, which were equipped with a Pt heater and interdigitated electrodes. The silicon substrates guarantee a sensor low-power consumption [3].
The gas sensors produced were first tested and calibrated in a sealed aluminum gas chamber, by using certified cylinders and mass flow controller to regulate the target gas concentration inside the test chamber. Gases analyzed in the laboratory were CO, NO2, NO, O3 and CH4, at different concentrations, based on the concentration limit defined by the EU law. The sensing responses were collected by thermo-activating the sensing materials at different temperatures.
The on-field sensing measurements were carried out in the city of Trento by using two sensor arrays developed in FBK. Each FBK sensor arrays were composed of eight MOX gas sensors, chosen based on the results obtained in the laboratory measurements. The two systems were also equipped with a commercial pump, to inject a constant air flux inside the sensors chamber, and a Raspberry Pi 3 model B, to transmit data via wireless to a dedicated FBK cloud. The two gas sensor arrays were placed in the APPA monitoring stations, in order to use the same sampling air analyzed by the certified instruments of the environmental protection agency.
Results and Conclusions
The laboratory measurements were carried out to identify the most promising gas sensors to be used in the arrays for the on-field analysis. For this purpose, the different sensors were exposed to various concentrations of pollutant gases, in the presence of 30%-60% of RH%. Cross-selectivity of the sensors was also tested to get as close as possible to the real on-field conditions, by injecting several target gases into the gas chamber at the same time. In Figure 1a) is reported the cross-selectivity characterization carried out by first injecting CO into the measuring chamber, and then a mix of CO and NO2. Figure 1b) shows some representative responses obtained with WO3, STN and SnO2 gas sensors. It can be observed that the presence of NO2 did not affect the STN sensing response vs. CO, while WO3 and SnO2 gas sensors resulted to be more reactive vs. NO2 than CO. The laboratory calibration allowed to identify the four best sensors for our aim, i.e. WO3, ZnO nanograin, SnO2 and STN.
In Figure 2a) and b) are reported the two gas sensor arrays developed. To verify the repeatability of the sensors' behavior, it was decided to use 2 WO3, 2 ZnO, 2 SnO2 and 2 STN sensors for each gas sensor array. The two systems were placed in two different APPA monitoring stations in Trento, to measure both urban background and traffic-oriented locations. The on-field measurements, which began on 27 August 2019, are still ongoing. Figure 3 shows the sensor signals collected in a month at the monitoring station for the detection of urban background pollution. The calibrations of the sensors carried out in the laboratory have been expanded by a further calibration on-field, by combining the sensor signals and data from certified monitoring systems with machine learning. A preliminary result, presented in Figure 4, shows the comparison between the daily average CO concentration determined by the certified sensor and that measured by our array of sensors, over a period of 2 weeks. In the next few months, the validity of the two sensor arrays for the detection of the most common pollutants will be investigated.
References
[1] https://unfccc.int/process-and-meetings/the-paris-agreement/the-paris-agreement
[2] Liu, X., Cheng, S., Liu, H., Hu, S., Zhang, D., Ning, H. A survey on gas sensing technology (2012) Sensors (Switzerland), 12 (7), pp. 9635-9665. doi: 10.3390/s120709635
[3] Bagolini, A., Gaiardo, A., Crivellari, M., Demenev, E., Bartali, R., Picciotto, A., Valt, M., Ficorella, F., Guidi, V., Bellutti, P. Development of MEMS MOS gas sensors with CMOS compatible PECVD inter-metal passivation (2019) Sensors and Actuators, B: Chemical, 292, pp. 225-232. doi: 10.1016/j.snb.2019.04.116
Figure 1