Date of Award

Spring 5-2018

Level of Access Assigned by Author

Open-Access Dissertation

Language

English

Degree Name

Doctor of Philosophy (PhD)

Department

Interdisciplinary Program

Advisor

John Vetelino

Second Committee Member

Brian Frederick

Third Committee Member

Robert Lad

Additional Committee Members

David Frankel

Guido Faglia

Abstract

Chemiresistive metal oxide gas sensors based on materials such as SnO2, ZnO, and TiO2, have been investigated extensively by many researchers for a wide range of applications. The band bending model, based on the surface chemistry of highly reactive ionosorbed species (O2- or O-) and the semiconducting material properties of SnO2, TiO2 and ZnO, adequately predicts the dependence of the change in sensor conductivity (Δσ) as a function of target gas pressure and temperature. However, the band bending model is not applicable to gas sensors based on reducible oxides such as WO3, MoO3 and V2O5, in which lattice oxygen reacts with adsorbed target gases creating oxygen vacancies which diffuse rapidly into the film bulk. In this study, a bulk conduction model which includes the kinetics of surface reaction and vacancy diffusion is developed in order to predict the response of reducible oxides,exposed to a target gas. In particular, , is predicted to be independent of film thickness, in contrast to the band bending model in which is a function of film thickness. The time dependence of the sensor response was also analyzed and it shows that the response time increases with film thickness in the bulk conduction model. In contrast, the response time for the band bending model is independent of film thickness. In order to validate the theoretical predictions of the bulk conduction model, experimental data was obtained by exposing ppb levels of NH3 gas to glancing angle deposited (GLAD) film with nanorod morphology and planar Au/WO3 film sensors at 450 °C. Δσ as a function of NH3 concentration was linear with a slope of unity for all the films and was also independent of volume to area ratio. Since the bulk conduction model predicts Δσ to be linear with a slope of unity with target gas concentration in the low pressure regime and also predicts Δσ to be independent of volume to area ratio, there is good agreement between the model and the experimental sensor response. This clearly indicates that the sensor response is due to bulk conduction mechanism. In summary it has been shown that theoretical model for bulk conduction mechanism is most appropriate to describe the response of WO3 film gas sensors as opposed to the band bending model.

Comments

Interdisciplinary in Sensor Science and Engineering

Download

Share