4.1 Single Micro PDMS PEMFC
4.1.2 Effect of Current Collector Shape
In this part, 5 current collector shapes under the same open ratio of 50% were designed, which are 1-circle, 4-circle, 9-circle, 16- circle, and 25-circle, as shown in Fig. 3.15. Unlike the hydrogen in anode flowing through the channel on the flow field plate, in the air-breathing cell, air in
67
cathode has to diffuse from the environment into the cell by itself without any flow field plate. And then the air diffuse further to the membrane through the holes on the current collector, releasing electrons from the GDL to the current collector, and meanwhile forming water on the membrane.
As shown in Figs. 4.3 and 4.4, which are I-V and corresponding I-P curves under different current collector shapes, even under the same open ratio, the performance increases with an increase of the circle numbers from one to twenty-five. That means the better reactions and more electrons are achieved for the cell having more holes on the current collector. Two possible reasons contribute to this result: one is that the average distance the electrons have to pass from the GDL to the current collector becomes shorter; the other is that the non-open area blocking the fuel diffusion becomes smaller; see Fig. 4.5.
When the chemical reaction is processing, electrons are not only produced on the area covered by the current collector, but also on the open area. If the electrons are produced on the open area, they have to pass through the GDL to the current collector, meaning that the path is with higher electric resistance such that lowers down the cell performance.
Besides, the gas and water produced have to remove and transport by diffusion between the covered and uncovered area of current collector. If the covered part is too deep for the fuel to diffuse, the performance will also decrease because of the difficult diffusion.
However, the membrane would become more humidified if the water could not diffuse smoothly and be carried away by the gas, then the total cell resistance would be lowered. Therefore, from the I-R curve
68
shown in Fig. 4.6 it is found that, although the total cell resistance decreases from 1-circle one to 25-circles one, making the maximum current density increase more, the cell resistance of the 16-circle is smaller than that of the 25-circle one because of the better water diffusion effect.
Table 4.2 summarizes the experimental results of the peak power density, the increase ratio of peak power with respect to the one of reference case, and the maximum current density that each cell can be extended for five different current collectors for the air-breathing cell (Figures 4.3, 4.4 and 4.6). It can be seen that both the Peak Power Density and Max Current Density Extended at 0.3V increase with the increment of hole number on current collector. However, the peak power increase ratios in 16- and 25-circle cases are 4.3% and 11.5%
comparing to that of reference case. Therefore, considering the factors among of structural strength, ease of manufacture, and increasing percentage of performance, this study proposes 9-circle current collector to be used in the present design for both air-breathing and forced convection cells. This is also the reason why this work selects the 9-circle current collector in the reference case.
69
Table 4.2 Experimental Results of Five Different Current Collector for Air-breathing Cell 25-circle 144.7 11.5% 370.5 4.1.3 Effect of Convection Type
Different from the air-breathing cell, the air is supplied through the flow channel on cathode in the forced convection one. From Figures 4.7 and 4.8, which are I-V and the corresponding I-P curves under forced convection, it can be found that in the low current density region, the I-V curve of 25-circle is below those of 9-circle and 16-circle at I <
162mA/cm2 and I < 261 mA/cm2, respectively that is different those in Fig. 4.2 for air-breathing cell. It is because forced convection tends to carry away the water and make the membrane dryer. Just as mentioned previously in air-breathing cell, the current collector with fewer holes can block more water from diffusing away between the current collector and GDL, so at lower current density region, where the electric chemical reaction is not too strong and produces less water, the membrane becomes dryer in 25-circle one that makes it perform worse than the other two.
Behind those two critical current densities, the performance of 25-circle current collector becomes better because the reaction is stronger and membrane is wetter. Even though, the performance trend of the forced convection cell is quite similar to that of air-breathing single cell at high
70
current density region. So does the I-R curve as shown in Fig. 4.9. In other words, the qualitative difference in performance among these shapes of current collectors is not affected by the convection way. In addition, Figure 4.9 (I-R curves) shows that the cell resistance of the 16-circle is almost coincident with the one of 25-circle because of the humidify effect mentioned previously in the air-breathing cell.
Now, compare the quantitative difference between the air-breathing and forced convection cells by using the 9-circle current collector (see the reference case in Table 4.1). As shown in Figs. 4.10 (I-V curve) and 4.11 (I-P curve), there exist intersections (occurred at I = 200 mA/cm2 approximately) on both I-V and I-P curves for these two types of cells. It is explained that although forced convection can enhance the air diffusion rate, it will accelerate the heat loss to lower down the temperature, which the electrochemical reaction needs. Besides, the water evaporation rate is simultaneously increased so that the water produced by electrochemical reaction is insufficient relatively. The combined effects of the temperature drop and higher water evaporation rate result in the resistance of electrochemical reaction and ion conductivity of membrane, causing the performance drop accordingly at lower current density region. Behind that (I > 200 mA/cm2 ), concentration losses for forced convection cell is lower owning to the sufficient fuel diffusion rate helping to eliminate the concentration loss, resulting in a better performance than that of air-breathing cell.
From above discussion, it can conclude that forced convection cell is a better choice for long-time high current-density output; in contrast, air-breathing cell is more suitable for lower current density output.
71
4.1.4 Clamping Force Effect
4.1.4.1 Micro Air-breathing PDMS PEMFC
In this part, the clamping force effect on the performance of air-breathing cell is observed. The torque on each bolt on the cell (4 bolts total) was increasing from 0.5 to 3.0Kgf．cm. The results are shown in Figs. 4.12 (I-V curve) and 4.13(I-P curve). It can be seen that the performance of the cell has a significant improvement when the clamping torque of each bolt is raised from 0.5 to 1.0Kgf．cm, increases slowly from 1.0 to 2.5Kgf‧cm, and starts to decrease after 3.0Kgf．cm.
From the I-R curves shown in Fig. 4.14, however, the cell resistance is all decreased as the clamping torques rising from 0.5 to 3.0Kgf‧cm.
Basically, the cell’s performance is improved by the applied assembly pressure because of the reduction in the electrical contact resistance between the current collectors and the GDL. However, when the applied pressure is getting higher, the porosity of the GDL and the size of the fuel flow channel will be squeezed simultaneously, causing the resistance increase of the fuel diffusion. The total performance will not be improved if the resistance of fuel diffusion is too much to be compensated by the reduction of the electrical contact resistance.
Therefore, an appropriate clamping torque should be considered carefully to enhance the performance without damaging the GDL and narrowing down the fuel flow channels. In the present study, a torque of 2.5 Kgf‧cm is recommended
72
4.1.4.2 Comparison of PDMS- and Silicon-Base Cells
As mentioned in sec. 1.1, since the structure of silicon-based flow field plate is too weak to prevent the problem of crack while assembling, the largest clamping torque applied can only be 0.5Kgf．cm. Take PDMS- and Silicon-Base cells for comparison under the same clamping torque of 0.5Kgf．cm, whose results are shown in Fig. 4.15 (I-P and I-R curves).
It is found that no matter in the power or the resistance exhibitions, the PDMS-based PEMFC performed much better than the Silicon-based one.
Their reasons are as follows: First, silicon is not as elastic as PDMS so that it cannot make the perfect contact between the current collectors and the carbon fiber on the GDL, causing the higher resistance; Second, silicon’s nonelastic property also produces the gas-leaking problem that lowers down the performance; Third, the non-opaque property of Silicon makes the assembly hard to be aligned precisely between constituting parts that may bring about more unpredictable problems. Owing to the elastic and transparent property, PDMS-base cell apparently is more suitable for the micro air-breathing PEMFC than the silicon-based one. It can also be seen from the I-R curve in Fig. 4.15 that the resistance of the PDMS-base cell is smaller than that of silicon-base cell. The reasons are similar to what just discussed above.
4.1.5 Durability Test of Single Micro PDMS PEMFC
To investigate whether the micro PDMS PEMFC performance is still able to be stable after the long-time test, the single micro PDMS PEMFC was tested for 14 hours continuously at different fixed operating voltages,
73
which are 0.7V, 0.6V, 0.5V and 0.4V, respectively.
Figure 4.16 shows the results at four different operating voltages, which are presented by power density as a function of time. In the first two hours, the four performance curves appear to have a little instability because the temperature, humidity and water management of fuel cell do not reach balance conditions yet; however, the system becomes stable obviously after the two hours. This result means that the single micro PDMS fuel cell can maintain a stable power output for a long time use up to 14 hours.
4.2 Micro Planar Air-breathing PEMFC Stack
In this part, a stack made of four micro PDMS planar air-breathing PEMFCs in series was designed; see Fig. 3.19, which used 10 bolts around four cells, and the 9-circles current collectors were adopted on both anode and cathode.
4.2.1 Effect of Clamping Force
Just like the single cell, increasing the clamping torque reduces the electric contact resistance and enhances the performance, but still there is limitation. In this experiment, the fuel flow rate was kept 60sccm in anode, and air-breathing was used in cathode. As shown in Figs. 4.17, 4.18 and 4.19, which are I-V, I-P and I-P curves, the range of the power increase is getting smaller as the applied torque becomes greater, implying that the assembly force cannot be over a certain limit, or the performance will not be further improved but damages of the GDL and flow channels can be possibly caused.
74
4.2.2 Effect of Fuel Supply condition
In this part, effects of two types of hydrogen supply conditions are studied. One is that the outlet is dead-end with a constant gage pressure of 0.5atm applied and kept at inlet; the other one is that both inlet and outlet are open with H2 supply volume flow rates of 60sccm and 120sccm, respectively. Dead-end means the outlet of the channels is closed that fuel must be used up within the cell. In this experiment, air-breathing was also used as the air supply in cathode, and the applied clamping torque was kept 1.5 Kgf．cm.
4.2.2.1 Effect on the Whole Stack
Figure 4.20 shows the I-V curves. For open outlet condition, it can be seen that, although all the I-V curves are almost coincident; however, the performance of 40sccm is better than those of 60 and 120sccm, especially in the larger current density range. Apparently, the air-breathing way cannot provide enough air into stack for the reactions of 60 and 120sccm of H2 supply. Such oversupply of H2 (≥60 sccm) leads to a lower reaction, similar to the situation of fuel-rich combustion.
In dead-end condition, the performance at low current density is almost the same as that of the open outlet condition with a flow rate of 60sccm, and a litter lower than that of 40sccm, moreover, the limiting current density can be extended to a higher value (from I = 1103mA to I = 1209mA), implying that the concentration loss is significantly improved due to the reinforced complete reaction in that region.
Obviously, the increase of flow rate is not a best way to improve the concentration loss in this air-breathing stack experiment because of the
75
limited air supply.
4.2.2.2 Effect on Each Cell
Designate the cell connecting to hydrogen inlet as the first cell and the one to outlet as the forth cell. From I-V curves for each cell shown in Figs. 4. 21, 4.22 and 4.23 under different fuel supply conditions, it can be found that, no matter what kind of fuel supply condition, the third cell always has the best performance, and the second one is next, after that is the first one, and the forth has the worst performance. As expected, the forth cell has the worst performance because this flow field design cannot supply the enough fuel in time to avoid the serious concentration loss at the end. And the third cell, which has the best performance, is mainly affected by the temperature and humidity. When the fuel travels through the first cell to the third one, the temperature has already been reheat from room temperature to higher one. Similarly, the fuel has also been humidified a lot that lowers down the resistance. These two factors make the electric chemical reaction to be more favorable. Therefore, owing to the temperature and the concentration loss effects, each cell in the same stack has different performance under this design (in series arrangement).
4.2.3 Stack Durability Test
In the durability test of the micro planar PDMS PEMFC fuel cell stack (in series arrangement), it was tested at various fixed operating voltages which are 2.8V, 2.4V and 2.0V (averaged 0.7V, 0.6V, and 0.5V for each cell), respectively. The results are also presented by power density as a function of time in Fig. 4.24. Similar to the single cell durability test, in the first two hours, these thee stack performance curves
76
appear to have a little instability and the system becomes stable obviously after the two hours. These results mean that the stacks can maintain a stable power output for a long time use up to 14 hours approximately.
4.3 Comparison between Silicon Base and PDMS Base
4.3.1 Forced Oxygen Supply
To compare the performance between the silicon-based flow field plate and the PDMS-based one, one result in Cheng’s study [1] was taken to make a comparison with the present one of PDMS under the same experimental parameters, listed on Table 4.3.
Table 4.3 Testing Conditions Reactant Gases
Gas reheat temperature 60℃
Gas humidified temperature 60℃
Open Ratio of Flow Field Plates Anode Open Ratio 75%
Cathode Open Ratio 75%
As mentioned previously, the present new design and fabrication method not only make assembly more convenient and mitigate the crack and gas-leakage problems, but also help the micro PEMFC planar stack more easily produced. Because the problems mentioned above are improved, it is not surprising that the PDMS-based one has a better
77
performance with a 24% of peak power more than that of the silicon-based one as shown in Fig. 4.25.
4.3.2 Air-breathing
Similar to last section, the air-breathing cell performance now is compared between Chang’s and present works. The experimental parameters are listed in table 4.4.
Table 4.4 Testing Conditions of Case Reactant Gases
Anode H2 (99.9%)
Cathode Air Flow Rates
Anode H2:30sccm ( Forced convection ) Cathode Air-breathing ( Natural convection )
Gas Backpressure (gauge)
Anode 0 kPa
Current Collector Slices
Material Cu/Au Temperatures
Gas reheat temperature Room Temperature(25℃) humidity Room condition (40%)
Open Ratio of Flow Field Plates Anode Open Ratio 75%
Cathode Open Ratio 75%
Figure 4.26 is the I-P curves for both experimental studies according to the test conditions of Table 4.4. Similarly, the PDMS-based one has a better performance than that of the silicon-based one with 24.5% more in peak power. It also means that the new design and fabrication method of PDMS PEMFC can improve the performance even for the air-breathing is
78
the way of air supply.
4.3.3 Liquid Water Produced by Thermal Effect
To observe the phenomena of liquid water produced on the GDL and flow field plate, both silicon-base air-breathing cell and the PDMS-base one were taken apart after they were tested for 15 minutes at a fixed voltage of 0.5V with the parameters the same as those in sec 4.3.2. The pictures are shown in Fig. 4.27 and 4.28, respectively. It can be seen that there are some liquid water forming on the silicon-base cell’s GDL and flow field plate, whereas no water is found on the PDMS-base cell’s ones because trace-amount of water is evaporated quickly as soon as the cell is disassembled. The reason is that PDMS is a material with much lower heat conductivity than that of silicon, so it can prevent heat loss by conduction of the flow field plate during electric chemical reaction, and make the cell temperature rising quickly to effectively evaporate the liquid water into vapors, easily carried away by the gas flow.
The micro air-breathing PEMFC produces less heat than the non-micro one. If the heat is lost too much, then it will lower down the temperature further and make more water vapors condensed into liquid water, causing water flooding in the cell. In the future the micro air-breathing PEMFC is used in commercial goods, it is very important to avoid the water flooding problem that can possibly damage the 3C products. PDMS, with less water flooding effects, apparently is a better material than silicon for micro air-breathing fuel cell.
79
'No loss' votage of 1.23V
Concentration losses
Fig. 4.1 Performance Curves (I-V, I-P and I-R) for Reference Case
0 50 100 150 200 250 300 350
current density (mA/cm2)
AC resistance meter value of Voltage/current
Fig. 4.2 Resistance Value from the Voltage Divided by the Current
80
Fig. 4.3 I-V Curves of Five Different Current Collectors for An Air-breathing Cell
Fig. 4.4 I-P Curves of Five Different Current Collectors for An Air-breathing Cell
81
Fig. 4.5 Comparison of Average Distance of Open Area and Non-open Area among Difference Current Collectors
0 50 100 150 200 250 300 350 400
0.0 0.1 0.2 0.3 0.4
370.525 340 352
284.575 180
1 circle 4 circle 9 circle 16 circle 25 circle
Current density(mA/cm2) Resistence(Ohm.cm2 )
Fig. 4.6 I-R Curves of Five Different Current Collectors for An
Air-breathing Cell
82 Fig. 4.7 I-V Curves of Five Different Current Collectors under
Forced Convection
Fig. 4.8 I-P Curves of Five Different Current Collectors under Forced Convection
83
Fig. 4.9 I-R Curves of Five Different Current Collectors under Forced Convection
Fig. 4.10 I-V Comparison of Forced and Natural Convection Effect
84
Fig. 4.11 I-P and I-R Comparison of Forced and Natural Convection Effect
Fig. 4.12 I-V Curves of Clamping Fore Effect on Single Cell
0 50 100 150 200 250 300 350 400 450
85 Fig. 4.13 I-P Curves of Clamping Fore Effect on Single Cell
Fig. 4.14 I-R Curves of Clamping Fore Effect on Single Cell
0 50 100 150 200 250 300 350 400 450
86
Fig. 4.15 Comparison of PDMS and Silicon Base
0 2 4 6 8 10 12 14 16 18
Fig. 4.16 Long Time Test of Single Micro PDMS PEMFC
87
0 200 400 600 800 1000 1200
1.5
Fig. 4.17 I-V Curves of Clamping Force Effect on Micro Planar PEMFC Stack
0 200 400 600 800 1000 1200
0
Fig. 4.18 I-P Curves of Clamping Force Effect on Micro Planar
Fig. 4.18 I-P Curves of Clamping Force Effect on Micro Planar