to air (non-pyrophoric); (4) capability for regeneration of deactivated catalysts by simple air- atmosphere annealing; and (5) availability of conventional ...
CERIA-BASED WATER-GAS-SHIFT CATALYSTS S. Swartz, A-M. Azad, M. Seabaugh NexTech Materials, Ltd., Worthington, OH Introduction Proton-exchange membrane (PEM) fuel cells are being developed for automotive (motive and/or auxiliary) and stationary (residential) power applications. PEM fuel cells operate either on pure hydrogen or a hydrogen-rich gas with little or no carbon monoxide. In the near term, fuel cells will need to be operated using our existing hydrocarbon fuel infrastructure (e.g., natural gas, propane, gasoline or diesel). Fuel processors convert these hydrocarbons into a “clean” hydrogen fuel needed for PEM fuel cell operation. The water-gas-shift (WGS) reaction (CO + H2 → CO2 + H2O) is a critical reaction used in fuel processors. This reaction increases the hydrogen content and reduces the carbon monoxide concentration, prior to final CO “cleanup” in subsequent preferential oxidation (or methanation) steps. Commercial WGS catalysts, designed for steady-state operation, are unsuitable for fuel cell applications because of their tendency to degrade under transient (e.g., start-stop, load-following) conditions encountered real systems. Over the past few years, NexTech Materials has pursued the development of WGS catalysts based on ceria-supported precious metals. Our work has demonstrated that Pt/ceria catalysts offer several advantages compared to existing copper-based catalysts: (1) operation over a wider range of temperature (up to 400ºC); (2) no need for activation prior to use; (3) no degradation upon exposure to air (non-pyrophoric); (4) capability for regeneration of deactivated catalysts by simple airatmosphere annealing; and (5) availability of conventional washcoating technologies for supporting ceria-based catalysts on monoliths to improve ruggedness. This work has shown that performance levels achieved in these catalysts are sufficient for fuel processors, and that shift reactors of small size and weight, and more importantly low cost, can be produced using monolith-supported forms. A number of groups have reported on the WGS performance of ceria-supported precious metal catalysts, with widely varying results [1-5]. Deactivation of Pt/ceria catalysts has been identified as a critical issue, and attributed to a number of mechanisms: over-reduction of ceria, build-up of carbonaceous species on the catalyst, and loss of precious metal dispersion. Another identified issue is methane formation through the reaction (CO + 3H2 → CH4 + H2O) under conditions of high temperature and low steam contents [6]. NexTech’s research has focused on optimization of composition and synthesis conditions, with the aim of identifying stable, methane-resistant catalyst formulations. A number of promising formulations have been identified, as described in this paper. Results and Discussion NexTech’s micro-reactor system for WGS testing is shown in Figure 1. The system design utilizes commercially available membrane-based humidifiers and dehumidifiers. This provides stable system operation over hundreds of hours, as long as heated water is provided to the humidifier and nitrogen purge gas is provided to the dehumidifier. Pre-mixed simulated reformate test gases are delivered through a single MFC to the humidifier and then through electrically heated gas lines (1/8-inch stainless) to the reactor. The reactor section incorporates a bypass loop, which allows for baseline gas chromatograph readings to be taken on humidified (non-reacted) gas, for subsequent conversion calculations. The reactor itself is a 1/8-inch diameter chamber, accommodating a powdered catalyst sample of 50 to 150 mg, and is electrically heated and electronically controlled.
Gas Chromatograph
H2/He/CO/CO 2
Membrane Dehumidifier
MFC
Membrane Humidifier
Reactor
Bypass
Figure 1. Schematic of micro-reactor used for WGS activity testing. NexTech’s testing methodology involves isothermal testing at 250 to 300ºC over a time period of up to 100 hours to assess catalyst deactivation rates. After certain periods of time on stream, conversions are measured at different temperatures to assess Arrhenius behavior. Testing was conducted with using a gas composition consisting of 8% CO, 12%CO2, 32% H2, 31%He, and 17%H2O, and a total flow rate of about 240 cc/min, and a catalyst weight of 100 to 150 mg. Isothermal tests at 250 to 280ºC were conducted over durations of 50 to 100 hours to assess catalyst deactivation rates. After isothermal testing, conversions were measured at different temperatures to assess Arrhenius behavior. Results of these tests are presented and their implications are discussed below.
Tests were conducted using a commercial copper-based catalyst (Süd-Chemie, C18-7) under the same conditions being used for Pt/ceria catalysts. Data presented in Figure 2 compares long-term stability data at 250ºC for the C18-7 catalyst and a Pt/ceria catalyst sample. At this temperature, the C18-7 catalyst provides very stable performance after an initial break-in period. Conversely, the Pt/ceria catalysts exhibit much higher initial activity but deactivate at a higher rate, with a fairly rapid deactivation over the first ten hours followed by a slower deactivation with increasing time. As shown in Figure 3, commercial copper-based catalysts have a relatively low activation energy compared to Pt/ceria catalysts. Thus at the higher operating temperatures, Pt/ceria catalysts are considerably more active than copper-based WGS catalysts.
Deactivation of Pt/ceria WGS catalysts, as reported by others and shown by data in Figure 2, is an issue that needs to be addressed, either by adapting previously demonstrated regeneration schemes [5], or by optimizing catalyst formulations to reduce the deactivation rates. This latter approach is demonstrated by data presented in Figure 4, where WGS data are presented for Pt/ceria catalyst formulations with substantially lower deactivation rates. The Pt/ceria catalyst sample (B) shown in the figure was prepared with a formulation similar to the Pt/ceria sample (A) shown in Figure 2. Samples C and D have the same Pt content (1 wt%) but were modified with a proprietary dopant, or “promoter”. The enhanced stability shown by these promoted catalysts suggests a viable development path for non-pyrophoric catalysts with enhanced stability, and potentially will minimize the need for (or frequency of) catalyst regeneration treatments.
CO Conversion (%)
100
75 S ample (A) 50 S ud-Chemie (C18-7)
25
0 0
20
40
60
Time on S tream (hours) Figure 2. WGS activity at 250ºC for a commercial copper-based catalyst and a Pt/ceria (3 wt% Pt) catalyst, under identical testing conditions (T = 250ºC, SV ~ 144,000 cc/g-hr).
CO Conversion (%)
100 Equilibrium 75 S ample (A)
50
S ud-Chemie (C18-7) 25
0 200
240
280
320
360
o
Temperature ( C)
Figure 3. Temperature dependence of WGS performance for a commercial copper-based catalyst and a Pt/ceria (3 wt% Pt) catalyst, under identical testing conditions (SV ~ 144,000 cc/g-hr).
Selectivity of WGS catalysts toward methane formation can be a problem, especially at higher temperature and with reformate gases with low steam content [6]. NexTech has addressed this problem by modifying the ceria support composition. As shown by gas chromatography data in Figure 5, methane formation is completely suppressed in these modified Pt/ceria catalysts. Acknowledgments
This work was funded by the U.S. Department of Energy (Contracts DE-FC02-98EE50529 and DE-FC02-99EE50586) and by the Office of Naval Research (Contract N00014-03-M-0041). The authors are grateful for collaborations with Ray Gorte (University of Pennsylvania), Jon Wagner and Mike Balakos (Süd-Chemie), and Tom Flynn and Jon Budge (McDermott Technology, Inc.).
CO Conversion (%)
100 90 S ample (D) 80 70
S ample (C)
60
S ample (B)
50 0
40
80
120
Time on S tream (hours)
Figure 4. WGS activity versus time for three Pt/ceria (1 wt% Pt) catalysts: non-promoted (B) and promoted (C & D) samples, under identical testing conditions (T = 280ºC, SV ~ 144,000 cc/g-hr).
CO2 CO2 CO CO H2
H2 CH4
Unmodifie d (T = 350ºC)
modified (T = 400ºC)
Figure 5. Gas chromatograph traces of Pt/ceria (2 wt% Pt) catalysts measured under equilibrium conditions, showing the effect of modifying the ceria support composition on methane formation. Literature Cited [1] X. Wang, et al., “Deactivation mechanisms for Pd/ceria during the water-gas-shift reaction”, J. Catal., Vol. 212, No. 2, pp. 225-230, Dec. 2002. [2] S. Hillaire, et al., “A comparative study of water-gas-shift reaction over ceria supported metallic catalysts”, Appl. Catal-A.-Gen., Vol. 215, No. 1-2, pp. 271-278, July 2001. [3] A.F. Ghenciu, “Review of fuel processing catalysts for hydrogen production in PEM fuel cell systems”, Curr. Opinion Sol. St. & Mater. Sci., Vol. 6, No. 5, pp. 389-399, Oct. 2002. [4] J.M. Zalc, et al., “Are noble metal-based water-gas shift catalysts practical for automotive fuel processing?”, J. Catal., Vol. 206, No. 1, pp. 169-171, Feb. 2002. [5] S.L. Swartz, et al., Ceria-Based Water-Gas-Shift Catalysts, Proceedings of the 2002 Fuel Cell Seminar, pp. 587-590 (Palm Springs, CA, Nov. 2002). [6] McDermott Technology, Final Report, DOE Contract No. DE-FC02-99EE50586 (Dec. 2002).