Supply and geopolitical issues associated with crude oil based energy have prompted the need for alternative sources of energy. Polymer Electrolyte Membrane Fuel Cells (PEMFC) have been proposed as an efficient alternative clean source of energy. The hydrogen required to fuel the PEMFC will eventually come from renewable sources such as wind and solar power. In the interim until these technologies mature, hydrogen will be generated from fossil or biologically derived hydrocarbon fuels using a process collectively called fuel processing.

The water gas shift reaction is a critical step in the conversion of hydrocarbon fuel into PEMFC quality hydrogen. The nature of on-site hydrogen generation requires precious metal based catalysts and catalyst cost is a major issue. Recent research has discovered polymetallic combinations of platinum-rhenium and platinum-rhenium-molybdenum to provide higher activity for water gas shift than monometallic precious metal catalysts. While these catalysts appear promising for water gas shift, virtually no work has studied the stability of these catalysts in realistic conditions.

My work studied the deactivation of a trimetallic Pt/ReMo/ZrO₂-containing water gas shift catalyst in realistic reaction conditions. A parametric study of catalyst stability revealed that while this catalyst is very stable above 300°C, the catalyst becomes increasing more unstable as it is operated below 300°C. Below 300°C, the stability of the catalyst is sensitive to synthesis gas composition, where higher carbon monoxide concentrations or higher ratios of steam to hydrogen result in more rapid catalyst decay. The mechanism for deactivation is fully reversible by heating in a reductive gas, which indicates the mechanism involves oxidation of the catalyst surface. Experimental observations are consistent with higher carbon monoxide coverage on platinum occurring below 300°C reduces the availability of activated surface hydrogen required to maintain the active state of the rhenium promoter, which is oxidized and rendered inactive by steam contained in the feed synthesis gas. Polymetallic formulations of Pt/Re and Pt/ReMo are inherently unstable for operation below 300°C and are not a realistic pathway for higher activity catalysts for low-temperature water gas shift.

As part of my investigation, the water gas shift and deactivation kinetics for the Pt/ReMo/ZrO₂-containing catalyst were measured. This was combined with transport theory to create a computational model of the catalytic reactor capable of predicting changes in concentration and temperature as a function of both time and space. This reactor model was combined with water gas shift experiments to study the interaction between catalyst deactivation and reactor design. This work discovered that when the CO conversion is high, axial heat migration is significant enough to increase the reaction temperature in the front of the catalyst bed, which will slow the deactivation of the trimetallic catalyst.

As a second part of my work, computational reactor modeling was applied to the mechanism by which diesel fuel decomposes to synthesis gas during autothermal reforming. A simplified reaction mechanism was developed for tetradecane (C ₁₄H₃₀) to gain insight into the reaction sequences that govern liquid fuel reforming as well as the probable pathways for coke formation. This analysis determined the tetradecane reaction sequence starts with a combination of oxidation and cracking reactions. Unexpectedly, the reforming of tetradecane and the hydrocarbons that result from the cracking reaction are not dominant reactions. What reforming does occur is limited to small hydrocarbons produced during the reaction sequence. In lieu of steam reforming, the hydrogen yield observed in experiments is primarily the result of partial oxidation reactions and, to a lesser degree, the water gas shift reaction. While the catalyst serves to initiate the reaction mechanism, a mixture of gas phase and surface reactions occur throughout the reactor.