Author

Mark W. Smith

Date of Graduation

2015

Document Type

Dissertation

Degree Type

PhD

College

Statler College of Engineering and Mineral Resources

Department

Chemical and Biomedical Engineering

Committee Chair

John W Zondlo

Committee Co-Chair

David Berry

Committee Member

Edwin Kugler

Committee Member

Edward Sabolsky

Committee Member

Charter Stinespring

Abstract

Two objectives for the catalytic reforming of hydrocarbons to produce synthesis gas are investigated herein: (1) the effect of oxygen-conducting supports with partially substituted mixed-metal oxide catalysts, and (2) a segmented bed approach using different catalyst configurations. Excess carbon deposition was the primary cause of catalyst deactivation, and was the focus of the experiments for both objectives. The formation and characterization of deposited carbon was examined after reaction for one of the selected catalysts to determine the quantity and location of the carbon on the catalyst surface leading to deactivation.;A nickel-substituted barium hexaaluminate (BNHA), with the formula BaAl 11.6Ni0.4O18.8, and a Rh-substituted lanthanum zirconate pyrochlore (LCZR) with the formula La1.89Ca0.11 Zr1.89Rh0.11, were combined with two different doped ceria supports. These supports were gadolinium-doped ceria (GDC) and zirconium-doped ceria (ZDC). The active catalyst phases were combined with the supports in different ratios using different synthesis techniques. The catalysts were characterized using several different techniques and were tested under partial oxidation (POX) of n-tetradecane (TD), a diesel fuel surrogate. It was found that the presence of GDC and ZDC reduced the formation of carbon for both catalysts; the optimal ratio of catalyst to support was different for the hexaaluminate and the pyrochlore; a loading of 20 wt% of the pyrochlore with ZDC produced the most stable performance in the presence of common fuel contaminants (>50 h); and, the incipient wetness impregnation synthesis method of applying the active catalyst to the support produced more stable product yields than the catalyst prepared by a solid-state mixing technique.;Different hexaaluminate and pyrochlore catalysts were used in different configurations in a segmented bed approach. The first strategy was to promote the indirect reforming mechanism by placing a combustion catalyst in the reactor inlet, followed by a reforming catalyst. This approach demonstrated that BNHA can be used in the reactor inlet to promote combustion with 1 wt% Rh-substituted pyrochlore in the reactor outlet, but the combustion catalyst should fill less than 50% of the reactor. The second approach placed specific catalysts in regions of the reactor that have conditions in which they are less likely to deactivate. This showed the most benefit in the use of a sulfur-tolerant noble metal catalyst in the reactor outlet.;The carbon formation study was conducted on a 2 wt% Rh-substituted pyrochlore. POX of TD for various run times, followed by temperature programmed oxidation, revealed two different types of carbon deposits in the catalyst bed: carbon that burned off at relatively low temperature (LTC), and carbon that burned off at higher temperatures (HTC). The LTC reached a steady state level within two hours of reaction, and was determined not to lead to catalyst deactivation. The HTC continued to accumulate with time on stream. A mathematical expression was developed to predict the rate of formation of the HTC for a given set of reaction conditions (O/C = 1.25). This expression was modified from data from a test under different reaction conditions (O/C = 1.1) for one length of time, and was found to predict the carbon formation for a different run time within 3%.

Share

COinS