CHAPTER 4 RESULTS AND DISSCUSSION
4.3 Effect of Different Catalyst Components
In this case, the effects of using different anode catalysts, which are pure Pt and Pt alloy, respectively, on CO tolerance are investigated. In the literature [4], it indicates that Pt alloy anode catalyst has a better CO tolerance for fuel cell. In generally, the anode catalyst for the
commercial MEA usually is Pt alloy and the cathode one is always the pure Pt. Because the activity of oxygen is lower than that of hydrogen, so, it needs a higher loading of Pt to perform reaction with catalyst. The Gore’s commercial MEA is used as fuel cell sample in present CO poisoning study. The loadings of Pt alloy and pure Pt are 0.45 and 0.6 , respectively, in this MEA. In present study, it uses pure Pt in anode catalyst to perform poisoning test. However, there is no such design in the existent commercial MEA. Therefore, the present work switches the cathode to anode and vice versa. Now, the cathode catalyst is Pt alloy and the anode one is pure Pt. Under this circumstance, the original baseline performance is expected to be different from the present one due to the different loadings. Therefore, it definitely causes the difficult to compare the CO tolerances directly for these two samples.
However, we still try to find the relation and trend between these two cases from the results of CO poisoning tests with and without air-bleeding.
/ cm2
mg
The procedure of CO poisoning tests are complete the same as the previous two sections. In the transient tests, the cell current density fixes at 800 and the anode is fed by pure hydrogen in the first 5min.
then, it is changed to , where the concentration is specified as 52.7ppm. In the transient air-bleeding tests, the introducing timing of air is 35min.
/ cm2
mA
CO
H /
2CO
Fig. 4.20 shows the transient curves with different anode catalysts, such as Pt alloy and pure Pt. In Fig. 20, the cell voltage declines very quickly when the CO contains in the anode fuel. The initial cell voltage
with anode catalyst of Pt alloy is 0.701V and the one with pure Pt anode catalyst is 0.657V. After switch, the cathode catalyst loading becomes 0.45 , less than the original one, 0.6 . The reduced loading causes the lower baseline performance and initial potential (0.657V) in transient test. In the original case with Pt alloy anode catalyst, it will reach a steady state after 45min in the CO poisoning test and the cell voltage is decayed to 0.417V (59.5%). In the switched case with pure Pt anode catalyst, it declines to a stable poisoned value of 0.303V (46%) after 47min. It can observe that the Pt alloy anode catalyst has a better CO tolerance and can reduce the decay rate of cell voltage as shown in Fig. 4.20. Also, it can reduce the CO adsorbability to Pt surface and causes an oxidation reaction of CO to remove it from Pt catalyst. Although the Pt alloy catalyst can improve CO tolerance in fuel cell, the reduction of CO poisoning effect is not so significant.
Even using Pt alloy as anode catalyst, the cell performance has a substantial drop by 52.7ppm of CO poisoning, especially after a long duration poisoning. The anode fuel will be turned back to pure when the CO poisoning effect reaches to a steady state in transient experiment. From these transient curves in Fig. 4.20, it can observe that the cell voltage is recovered rapidly for both pure Pt and Pt alloy anode catalyst because purging pure can change the balanced poisoning state. The Pt alloy catalyst has faster cell voltage recovery rate after purging pure into anode fuel stream. In other words, Pt alloy catalyst can remove CO more effectively.
/ cm2
mA mA/ cm2
H
2H
2H
2Fig. 4.21 shows the steady polarization curves includes the baseline,
poisoned and recovery performances with different anode catalysts. In this figure, the performance of baseline polarization curve with pure Pt anode catalyst is only 70% of that with Pt alloy one. The reason has been given previously. In Fig. 4.21, it also can find that the power output of poisoned polarization curve is very small with pure Pt anode catalyst.
The cell current density is only 137 less than 10% of base performance (1744 ) when the cell voltage is 0.496V. The pure Pt catalyst cannot avoid the poisoning of CO because CO almost occupies the whole active sites on catalyst surface when CO poisoning effect reaches the steady state. The Pt alloy catalyst has a better CO tolerance comparing with pure Pt catalyst. However, it only can increase cell performance from 10% to 20% (437 ) with respect to the baseline datum (2280.8 ) at the cell voltage of 0.51V. There is an obvious recovery of cell performance after purging pure for 30min.
/ cm2
After that, it will discuss the CO tolerance by using air-bleeding technique with pure Pt anode catalyst. It has been verified that the air-bleeding technique can effectively improve fuel cell CO (52.7ppm) tolerance previously when the anode catalyst uses Pt alloy as shown in Fig. 4.8, 4.9.
Fig. 4.22 shows the transient curves with 52.7ppm of CO poisoning by using air-bleeding technique with pure Pt as anode catalyst. The procedure of this transient test is the same as that in the previous case.
In this test, the CO of 52.7ppm is introduced into anode fuel stream at the time of 5min and the air is injected into anode at the instant of 35min.
From these curves, it can observe that the cell voltage is recovered quickly as air is injected into anode fuel stream. With 7% of air-bleeding, the cell has the optimum CO tolerance and the cell voltage can be recovered to 0.649V (98.8%) with current density 800 in the transient curve. In Table 4.4, it makes a summary of the transient test results, which can serve as a comparison between the Pt alloy and pure Pt anode catalysts. It shows that the air-bleeding can increase CO tolerance and improve cell performance no matter what kind anode catalyst is (pure Pt or Pt alloy). With Pt alloy anode catalyst, the cell voltage has optimum recovery rate of 98% when 4% air-bleeding is introduced into anode. However, the recovery rate can be further increased to 98.8% with 7% air-bleeding when the anode catalyst is pure Pt. It is because the loading becomes higher (0.6 ) as pure Pt is used as anode catalyst, comparing with the original one (0.45 ), it causes more adsorption of with Pt and more desorption of CO from Pt surface. Note that in this table, the resultant poisoning rates are different for these transient tests after the specific poisoning duration (30min). It was explained in the last section. However, the discrepancies among the poisoning rates are larger by using pure Pt as anode catalyst. It can be found a trend that the poisoning rate becomes greater as the number of continuous transient poisoned tests increases.
The reason may be that the interval between two consecutive transient tests is too short and the catalyst features may have some changes after the repeat transient poisoning experiments. It is believed that the results of transient air-bleeding tests are not affected by this drawback.
/ cm2
Fig. 4.23 shows the steady polarization curves with different anode catalysts and the percentage of air-bleeding. Without air-bleeding, it is observed that the Pt alloy anode catalyst has a better CO tolerance than that of pure Pt catalyst. With air-bleeding, it can remarkably improve the CO tolerance for both Pt alloy and pure Pt anode catalysts.
CHAPTER5 CONCLUSIONS
This thesis consists of three parts. The first one is to experimentally investigate the transient CO (52.7ppm) poisoning test with fixed cell voltage and current density conditions, respectively. With fixed cell voltage to perform transient CO poisoning test, the anode potential is restricted to change. Therefore, the CO adsorption rates at anode catalyst with any voltage conditions reaches to the same steady-state values. So, the poisoned polarization curves under different cell voltages to do CO poisoning tests are alike. For the CO poisoning tests with fixed current density, using higher current density can increase cell CO tolerance.
The higher current density can promote the anode potential to achieve the onset potential of CO oxidation that causes CO removed from Pt surface.
The steady adsorption rate of catalyst is decreased following the rise of current density condition in the transient tests. Apparently, it can improve the cell performance, CO tolerance, by using this method.
The second one is to investigate the effects of air-bleeding with different introduced timing (3 and 30min) in the transient poisoning CO tests. The CO concentrations are 52.7, 25 and 10.1ppm, respectively.
With the air-bleeding ratios of 4, 3 and 1.5% applied to each CO concentration, the fuel cell can obtain the optimum CO tolerance when the air-bleeding introducing timing is 30min. The steady recovery polarization behaviors can reach to 94, 96 and 97% of baseline performance, respectively, under the same current density. With the
air-bleeding ratios of 3, 1.5 and periodic 1.5% applied to each CO concentration, the cell can obtain the optimum CO tolerance while the air-bleeding timing is 3min in the transient experiments. With 3min air-bleeding timing, the steady recovery polarization curves can achieve over 97% of baseline performance for each CO concentration under the same current density. In this situation, the effect of CO poisoning is almost disappeared with lower CO concentration condition (25 and 10.1ppm). The air-bleeding is able to improve the fuel cell CO tolerance and recovery poisoned performance no matter what the air introducing timing. CO absorbed on catalyst surface reacts with injected to carry out the oxidation reaction, and it is removed from Pt surface to recover cell performance. The air-bleeding timing is very important factor for the cell CO tolerance performance. The optimum ratio of air-bleeding is decreased and the cell performance recovery rate is increased when the air-bleeding timing is short (3min). The adsorption ratio of Pt surface is lowered as the CO poisoning duration is shorter. In this situation, the injected has more chance to be adsorbed on the catalyst surface and then increase the reaction rate of
O
and CO.Therefore, it costs a lower air-bleeding ratio to get better CO tolerance performance by a shorter air injecting timing.
O
2O
22
The third one is to investigate the effect on the cell performance by using different anode catalyst component (Pt alloy and pure Pt) to perform the CO poisoning tests with and without air-bleeding. Without air-bleeding, the Pt alloy anode catalyst has a better CO tolerance comparing to the pure Pt anode catalyst. With the pure Pt catalyst, it
cannot reduce the CO (52.7ppm) poisoning effect and the power output is less than 10% of baseline datum at the cell potential of 0.5V. However, it is insignificant to increase cell CO tolerance by using Pt alloy anode catalyst. With air-bleeding, it can increase CO tolerance effectively no matter what kind anode catalyst (pure Pt or Pt alloy) is used to perform CO poisoning tests.
Finally, there are some suggestions for the future extensions of the present experiment. It could determine the life of a commercial MEA when CO is contained in the anode fuel stream. Also, the effect of air bleeding on the MEA life would be interesting since oxygen could cause the damage of MEA due to the heat generated from its reaction with hydrogen and the change of catalyst properties. Eventually, it might determine the fuel cell life that anode fuel is from reformer gas, which contains CO, and then it is subjected to air-bleeding.
REFERENCE
1. W. Nogel, J. Lundquist, P. Ross, P. Stonehart, Electrochim. Acta 20, pp.79 (1975).
2. R.M.Q Mello, E.A. Ticianelli, Electrochim. Acta 42, pp.1031 (1997).
3. H. P. Dhar, L. G. Christner, A. K. Kush, and H. C. Maru, Nature of CO adsorption during oxidation in relation to modeling for CO poisoning of a fuel cell anode, Journal of The Electrochemical Society, 134, pp.3021 (1987).
H
24. Watanabe M. and S. Motoo, Journal of Electroanalytical Chemistry, 63, pp.97 (1975).
5. Christoffersen E., P. Liu, A. Ruban, H. L. Skriver and J. K. Noskov, Anode materials for low-temperature fuel cells: A density functional theory study, Journal of Catalysis, 199, pp.123 (2001).
6. Watanabe M., H. Igarashi and T. Fujino, Design of CO tolerance anode catalysts for polymer electrolyte fuel cell, Electrochemistry, 67, pp.1194-1196 (1999).
7. H. A. Gastiger, N. Markovic, P. N. Ross Jr., E.J. Cairns, CO Electroxidation on Well-Characterized Pt-Ru Alloys, Journal of Physical Chemistry, 98, pp.617 (1994).
8. Hongmei Yu, Zhongjun Hou, Baolian Yi, Zhiyin Lin, Composite anode for CO tolerance PEMFC, Journal of power sources, 105 pp.52-57 (2002).
9. Andrew T. Haug, Ralph E. White, John W. Weidner Wayne Huang, Steven Shi, Narender Rana, Stephan Grunow, Timothy C. Stoner and Alain E. Kaloyeros, Using Sputter Deposition to Increase CO Tolerance in a Proton-Exchange Membrane Fuel Cell, Journal of The Electrochemical Society, 149 (7) A868-A872 (2002).
10. S. J. Lee, S. Mukerjee, E. A. Ticianelli, J. McBreen, Electrocatalysis of CO tolerance in hydrogen oxidation reaction in PEM fuel cells, Electrochimica Acta, 44, 3283-3293 (1999).
11. G. A. Camara, E. A. Ticianelli,S. J. Lee and J. McBreen, The CO poisoning Mechanism of the Hydrogen Oxidation Reacting in Proton Exchange Membrane Fuel Cells, Journal of The Electrochemical Society, 149 (6), A748-A753 (2002).
12. Shimshon Gottesfeld and Judith Pafford, A new approach to the
problem of Carbon Monoxide Poisoning in fuel cells Operating at Low Temperatures, Journal of The Electrochemical Society, 135, 2651-2652 (1988).
13. Shimshon Gottesfeld, Preventing CO poisoning in fuel cells, United States Patent, NO. 4,910,099 Date: Mar. 20,1990.
14. Knights et al., Method for operating fuel cells on impure fuels, United States Patent, NO: US6,500,572 B2 Date: Dec. 31, 2002.
15. Mahesh Murthy, Manual Esayian, Alex Hobson, Steve MacKensie, Performance of a Polymer Electrolyte Membrane Fuel Cell Exposed to Transient CO Concentrations, Journal of The Electrochemical Society, 148 (10), A1141-1147 (2001).
16. Mahesh Murthy, Manual Esayian, Woo-kum Lee and J. W. Van Zee, The Effect of Temperature and Pressure on the Performance of a PEMFC exposed to Transient CO Concentrations, Journal of The Electrochemical Society, 150 (1), A29-A34 (2003).
17. Jingxin Zhang, Tony Thampan and Ravindra Datta, Influence of Anode Flow Rate and Cathode Oxygen Pressure on CO Poisoning of Proton Exchange Membrane Fuel Cells, Journal of The Electrochemical Society, 149 (6), A765-A772 (2002).
18. R. J. Bellows, E. Marucchi-Soos and R. P. Reynolds, The Mechanism of CO Mitigation in PEMFC using Dilute in the Anode Humidifier, Electrochemical and State Letters, 1 (2), 69-70 (1998).
2 2
O H
19. J. Divisek, H. F. Oetjen, V. Peinecke, V. M. Schmidt and U. Stimming, Components for PEM fuel cell systems using hydrogen and CO containing fuels, Electrochimica Acta, Vol 43, NO.24, pp.3811-3815 (1998).
20. L. P. Carrette, K. A. Friedrich, M. Huber and U. Stimming, Improvement of CO tolerance of proton exchange membrane fuel cells by a Pulsing technique, Physical Chemistry Chemical Physics, 3, 320-324 (2001).
21. R. W. Fox and A. T. McDonald, Introduction to Fluid Mechanics, John Wiley and Sons, Canada, 1994
22. S. J. Kline and F. Mcclintock, Describing Uncertainties in Single-Sample Experiments, Mechanical Engineering, vol. 75, pp.3-8, 1953.
23. R. J. Moffat, Contributions to the Theory of Single-Sample Uncertainty Analysis, Journal of Fluid Engineering, vol. 104,
pp.250-260, 1982.
24. R. S. Figliola and D. E. Beasley, Theory and Design for Mechanical Measurements, 2nd Ed., John Wiley and Sons, Canada, 1995.
25. J. P. Holman, Experimental Methods for Engineers, Ed., McGraw-Hill, New York, 1989.
5th
26. A. Weber, R. Darling, J. Meyers and J. Newman, Handbook of Fuel Cells, Volume2 Fundamentals and survey of systems, pp.49, John Wiley & Sons (2003).
Table 1.1 Summary of investigation of CO tolerance on PEM fuel cells:
Catalyst:
Author Investigation method Parameter Results Remarks
Watanabe M. et al.
(4)
Pt-Ru alloy catalyst CO Ru form RuOH, CO oxidized by RuOH toremove from catalyst Christoffersen E. et al.
(5)
Used density functional theory to calculatedual metal catalyst surface characteristics and the corresponding CO adsorptive power
Pt-Ru dual metal catalyst
CO adsorbs on Ru of Pt-Ru alloy catalyst Decrease CO adsorbability with Pt surface
Hougmei Yu et al.
(8)
Composite anode: inner and outer catalyst Inner: Pt, outer:Pt-Ru alloy
Outer catalyst improves CO tolerance Inner + outer catalyst has better performance to traditional Pt-Ru alloy catalyst
The ratio of inner and outer catalyst
Ralph E. White et al.
(9)
Sputter deposition Ru on Pt surface Ru to form CO filter, Pt + Ru filter catalyst100ppm CO, : 2%
O
2 With suitable -bleeding ratio, this has better CO toleranceO
2 Pt-Ru alloy replacedRu may better J. Mcbreen et al.
(10)
investigate CO and hydrogen reactionmechanism on three types of catalyst
Pt, PtSn and PtRu catalyst
CO adsorption step, form of CO bonded and onset potential of CO oxidation change with different catalyst
J. Mcbreen et al.
(11)
further analyzed these catalyst kinetic mechanisms by kinetic model analysisPt, PtSn and PtRu catalyst
both bridge- and linear-bonded adsorbed CO appear on catalyst, CO oxidation initiates at bridge-bonded, CO oxidation/adsorption steps change by different catalyst
Table 1.1 Summary of investigation of CO tolerance on PEM fuel cells: (continuity)
Air-bleeding:
Author Investigation method Parameter Results Remarks
Shimshon G. et al.
(12)
Inject oxidant of CO, oxygen-bleeding CO: 100ppmO
2: 2~5%Injection 2~5% oxygen could complete restored cell performance (100ppm CO)
First literature of air-bleeding Shimshon G. et al.
(13)
oxygen-bleeding CO: 100~1000ppm 2~6% oxygen substantially restore cellperformance (100~500ppm), exceed cause performance decreasing
O
2: 2~6%O
2Knights et al.
(14)
Air-bleeding system by using a sensor cell Periodic air-bleedingAir-bleeding timing:
sensor cell voltage lost than 100 mV
prevent from poisoning without wasting air, pulsed air-bleed would have better CO tolerance than constant
J. W. Van Zee et al.
(15)
Compare with different GDL, transient poisoning test, air-bleedingGDL: SSE, CARBEL CO: 500, 3000ppm Air-bleeding
With bleed: decay rate lower, recovery rate higher, 5% air improve CO tolerance effectively (500ppm)
GDL: cannot define which one is better
J. W. Van Zee et al.
(16)
Discuss cell temperature and cathode backpressure, transient testT:
70
οC , 90
οC
P: 101, 202KPa Air-bleeding
Increase T: decrease CO adsorbability and poisoning rate of forteen times
Increase cathode backpressure: may crossover to anode and decrease poisoning rate of four times
O
2Table 1.1 Summary of investigation of CO tolerance on PEM fuel cells: (continuity)
Others:
Author Investigation method Parameter Results Remarks
Ravindra Datta et al.
(17)
The effect of anode flow rate and cathode oxygen pressureCO: 100ppm, fixed current density
flow rate :poisoning rate , increasing cathode oxygen pressure would cause diffusion through the membrane and improve CO tolerance
↑ ↑ The thick of MEA
R. J. Bellows et al.
(18)
diluted in the anode humidifier, decompose into oxygen and water2
At 100 ppm CO, the cell performance would almost restore with 0.75% in the
L. P. Carrette et al.
(20)
pulsing technique to change anode potential Cathode: reference electrodethe catalyst surface was continuously cleaned and the loss of cell voltage was lowest
pulse height: adjust for different catalyst
Table 2.1 Fuel cell operation conditions.
Cell temperature 65℃
Humidification temperature Anode: 80℃, Cathode: 70℃
Backpressure Anode, Cathode: 1atm Fuel flow rates
H : 10.4 cc/min/Amps
2Transient conditions: Fix cell voltage: 0.5, 0.6 and 0.7V Fix current density: 600, 1000 and 1200mA/ cm2
CO concentration 52.7ppm
Case 2
Feed stream Anode: , + CO, + CO +
air-bleeding
H
2H
2H
2Cathode:
O
2Transient condition Fix current density: 800mA/ cm2 CO concentration 10.1, 25 and 52.7ppm
Air-bleeding timing 30 and 3min
Case 3
Feed stream Anode: , + CO, + CO +
air-bleeding
H
2H
2H
2Cathode:
O
2Transient condition Fix current density: 800mA/ cm2 CO concentration 52.7ppm
Anode catalyst Pt alloy and Pt Air-bleeding timing 30min
Table 3.1 Uncertainty of electronic load potential meter
Table 3.2 Uncertainty of electronic load current meter
Fluke digital meterTable 3.3 Uncertainty of anode MFC
Table 3.4 Uncertainty of cathode MFC
Standard valueTable 3.5 Uncertainty of air-bleeding MFC
Output Voltage Brooks MFC readvalue (sccm)
Table 3.6 Uncertainty of anode temperature controller
Standard value (℃) Measure value (℃) Uncertainty (%)
25 25 0 35 35 0 50 50 0 70 70 0
85 84 -1.17
95 94 -1.05
100 99 -1
Table 3.7 Uncertainty of cathode temperature controller
Standard value (℃) Measure value (℃) Uncertainty (%)
25 25 0 35 35 0 50 50 0 70 70 0 85 85 0
95 94 -1.05
100 99 -1
Table 3.8 Uncertainty of cell temperature controller
Standard value (℃) Measure value (℃) Uncertainty (%)
25 25 0 35 35 0 50 50 0 70 70 0 85 85 0 95 95 0 100 100 0
Table 3.9 The table of experimental repeatability for baseline
Table 3.10 The table of experimental repeatability for baseline
Table 3.11 The table of experimental repeatability for baseline performance (case 3, Pt anode catalyst)
Cell
Table 4.1 Summary data from the transient CO poisoning experiments with two different specific times for 52.7ppmCO
52.7ppm CO, Specific time: 30min
Vref (V) Vspe (V) Air (%)
V
re (V)52.7ppm CO, Specific time 3min
Vref (V) Vspe (V) Air (%)
V
re (V)Table 4.2 Summary data from the transient CO poisoning experiments with two different specific times for 25ppmCO
25ppm CO, Specific time: 30min
Vref (V) Vspe (V) Air (%)
V
re (V)25ppm CO, Specific time: 3min
Vref (V) Vspe (V) Air (%)
V
re (V)Table 4.3 summary data from the transient CO poisoning experiments with two different specific times for 10.1ppmCO
10.1ppm CO, Specific time: 30min
Vref (V) Vspe (V) Air (%)
V
re (V)10.1ppm CO, Specific time: 3min
Vref (V) Vspe (V) Air (%)
V
re (V)Table 4.4 summary data from the transient CO poisoning experiments
with two different anode catalysts for 52.7ppmCO52.7ppm CO, Pt alloy
Vref (V) Vspe (V) Air (%)
V
re (V)Fig. 1.1 Scheme diagram of the thesis
Fig 2.1 Schematic drawing of overall experimental system
Fig. 2.2 The HP 6060B electronic load
Fig. 2.3 The MFC readout power supply
Fig. 2.4 The power supply of the test station
Fig. 2.5 The gas pipelines controller