Ph.D Thesis

Ph.D StudentOsnat Landau
SubjectNanostructured Titanium Oxide Gas Sensors Produced by
DepartmentDepartment of Materials Science and Engineering
Supervisor Professor Rothschild Avner
Full Thesis textFull thesis text - English Version


Nanofiber gas sensors produced by electrospinning have demonstrated ultrahigh sensitivity to sub ppm gas concentrations, short response and recovery time, good reversibility, reproducibility and stability. However, there are still unexplored areas and open questions that should be investigated in order to advance this dynamic research field and develop even better devices. The primary goal of this research was to investigate the correlation between the microstructure and gas sensitivity of porous TiO2 layers serving as a model system for metal oxide gas sensors. We focused on the effect of morphological features such as the grain size, porosity and specific surface area on the sensitivity, with a specific aim to compare fibrous vs. non-fibrous layers. Towards this end we produced fibrous and non-fibrous layers by electrospinning of polymeric solution comprising titanium isopropoxide and acetic acid, followed by hot pressing and calcination. The microstructure and chemical composition of the TiO2 layers were carefully characterized following the different steps in the fabrication process. The sensing properties of selected sensors with different microstructures and morphologies were examined upon exposure to traces of reducing (CO) and oxidizing (NO2) gases in air. The different response patterns displayed by these sensors were compared and correlated with their microstructure. The conclusions obtained from the empirical observations were supported by numerical simulations of the response dynamics. Our TiO2 layers displayed remarkable sensitivity. The responses were reversible with relatively short response and recovery time. Despite having distinctly different microstructures, the sensors displayed quite similar gas sensitivity with no significant variations between different sensors. This unexpected result led us to revisit our initial hypothesis that the sensitivity should always increase with increasing surface-to-volume (S/V) ratio and with higher porosity. To explain this observation we invoked a new mechanism, so-called the self-gettering effect that suppresses the sensitivity for sensors with excessively high S/V ratio. This effect counterbalances the grain size effect that is known to enhance the sensitivity in nanosized porous layers, resulting in an optimal morphology at which the gas sensitivity is maximal. In addition, we proved that the gas diffusion through our layers was clearly not the rate limiting step. We showed that the kinetics of chemical reactions involved in the gas sensing mechanism, rather than the gas transport kinetics, controls the response dynamics of our sensors. Taking these factors into account we were able to simulate the different response vs. time profiles observed in our gas sensing tests.