Development and Characterisation of SiC Based Sensors for Harsh Environments

Abstract: The introduction of chemical gas sensors in our environment is a process that started many years ago. However, although the process has been rather slow over the years, the interest for chemical gas sensors is rising due to the growing environmental awareness. The need to measure gas components in our surrounding is becoming more and more important as the technological level increases in the world.The development of gas sensors for different type of combustion control has been the focus in this thesis. To be able to measure in harsh environments such as flue gases or diesel exhausts, extremely robust sensors have to be developed. By using silicon carbide, a wide band gap material and chemically inert material, sensors that can withstand the harsh environment was developed and manufactured. Since SiC is a rather new material in the world of sensors, a great deal of work was performed on the design and basic understanding of the device. Several different ohmic contacts were studied thoroughly and it was shown that the ohmic contacts on the device could be improved. The effect of hydrogen diffusion into a SiC gas sensor device has also been investigated.Much effort was spent on optimisation of the design and measurement parameters for the sensor when detecting different gas components. The device design parameters are the choice of catalytic metal for the sensor, the thickness of the metal and the porosity of the metal. The operating parameters are the sensor temperature and the annealing procedure before use. It was shown that different pre-treatment of the sensor could have large influence on the ammonia response.Selected gas sensors were used in the flue gas project where the intention is to use the information about the gas concentrations to regulate the boiler for a better power-economy. SiC based Schottky diodes or transistors combined with commercial sensors and a linear lambda sensor was used in a high temperature electronic nose to measure in the flue gases. By using multivariate methods it was possible to evaluate the large amount of data from the continuous measurements in the flue gases. The results showed that it might be possible to use a small array of different gas sensors to identify different modes of operation of the boiler. Furthermore, predictions of the concentrations of a few specific gas components in the flue gases could be made.