ASSESSMENT OF PHOTOCATALYTIC PERFORMANCE OF ALUMINUM-SUPPORTED TITANIUM DIOXIDE THIN FILMS PRODUCED BY FLAME ASSISTED CHEMICAL VAPOUR DEPOSITION
Potter, Eric E.
MetadataShow full item record
Titanium dioxide films were deposited on aluminum by flame assisted chemical vapour deposition. The photocatalytic performance of these films when the air flow rate to the flame and thickness of silicon oxides under-layer were varied, was compared against a film of electrophoretically deposited Degussa P25, a titanium dioxide powder. At a high air flow rate, only powder was deposited on the substrate and the films had no measurable photocatalytic activity. When the air flow rate was lower, and the flame cooler, titanium dioxide was successfully deposited. Samples with a silicon oxides under-layer had more titanium dioxide deposited than with no undercoating, but titanium dioxide content did not increase as the layer of silicon oxides thickness increased. Films deposited at a low air flow rate, and with a silicon oxides under-layer, were shown to act as self-cleaning surfaces through the mineralization of a stearic acid film, successfully removed organic compounds from water through the oxidation of methylene blue as a model compound, and effectively inactivated Vibrio fischeri in water. However, in all cases the performance was found to be moderate when compared to Degussa P25 films. The successful use of films produced under these conditions in various applications suggests that they may also be effective in an air sterilization unit. An air sterilization reactor system was designed with considerations to reactor geometry, relative humidity and air flow rate. Bioaerosol generation and capture experiments were performed while varying pressure supplied to a Collison Nebulizer, the nebulizer optical density of Bacillus subtilis, and the sample period using a modified AGI-4 impinger. A model was developed to correlate these parameters to expected bacterial capture. Different methods of assessing cell viability in the capture solution were examined, and the use of the Live/Dead assay was recommended over traditional plate counts because of lower associated error.