Author

Donald Bragg

Date of Award

2007

Level of Access

Open-Access Thesis

Degree Name

Master of Science in Mechanical Engineering (MSME)

Department

Mechanical Engineering

Advisor

Howard H. Patterson

Second Committee Member

Michael T. Boyle

Third Committee Member

Justin Poland

Abstract

CO contamination is common in hydrogen extraction processes. Proton exchange membrane fuel cells need a source of hydrogen containing less than 10ppm of CO to avoid the irreversible poisoning of the fuel cell’s platinum anode. Platinum metal is commonly used as a catalyst for the purification of hydrogen gas. An effort has been put forth to replace the expensive platinum catalyst with a less expensive alternative. Photocatalytically active metal oxide semiconductors have shown promise as less expensive alternatives to platinum for redox reactions such as the oxidation of CO. This research studies a MoO3 sputter deposited thin film on a SiO2 substrate (MoO3/SiO2) for use as a photocatalyst for the oxidation of CO. The MoO3/SiO2 catalyst was run under atmospheric pressure at a temperature of 293K. The atmosphere used for all experiments was 900 ppm CO, 1800 ppm O2 with an Ar balance. The oxidation of CO occurs on the surface of the metal oxide catalyst and proceeds as follows: 2 CO + O ? CO The proposed mechanism for the oxidation of CO using photocatalytically activated MoO3/SiO2 catalysts is shown below. * 3 5 3 Mo6 O Mo O + h + ??? ? 2 2 * 4 3 Mo5 O + CO ???Mo O + CO + + 3 6 2 2 4 Mo O O(SiO ) Mo O + + + ??? Preliminary experiments utilized a chemically impregnated MoO3/SiO2 powder sample which demonstrated poor catalytic performance in the oxidation of CO. The impregnated catalyst turned a deep blue after illumination, indicative of the formation of MoO2, showing that the catalyst’s oxygen transport pathway was blocked, thus demonstrating the importance of the interface between the metal oxide and substrate material. Three thin film MoO3/SiO2 catalyst samples were prepared using sputter deposition to study the effects of varying the area of the MoO3/SiO2 interface. The samples consisted of: a monolayer sample, having the largest MoO3/SiO2 interface area; a bi-layer sample having one half the MoO3/SiO2 interface area of the monolayer sample, and a 60 Å thick sample which had the MoO3/SiO2 interface completely blocked. XPS was used to characterize the catalyst samples. The monolayer MoO3/SiO2 thin film catalyst had a catalytic efficiency resulting in 55% CO oxidation. The reduction in area of the MoO3/SiO2 interface has detrimental effects on the CO conversion of the catalyst. The bi-layer sample and 60 Å sample had CO conversions of 29.3% and 22.3% respectively. The proposed catalytic reaction mechanism for this system was supported by the experimental results showing that the MoO3/SiO2 interface was crucial for catalytic activity. The catalytic efficiency of the monolayer sample degraded with repeated use. Additional catalysts using a Si3N4 ion barrier coating placed between the SiO2 substrate and MoO3 thin film were prepared in order to explore this phenomenon. XPS analysis of the Si3N4 catalyst after experimentation revealed that the MoO3 thin film was still present; therefore, the MoO3 thin film on the original catalyst had migrated into the substrate as a result of pretreatment heating. A gas mixing system, catalyst pre-treatment system and photocatalytic reactor were designed and built. The detection method was GC/MS, to monitor O2 and CO2. The EntryRAE™ multi-gas detector was used to monitor CO and to prepare the CO, O2 and Ar gas mixtures.

Share