O2 Planar Membraneless Microchannel Fuel

3544

Langmuir 2005, 21, 3544-3550

A Dual Electrolyte H2/O2 Planar Membraneless Microchannel Fuel Cell System with Open Circuit Potentials in Excess of 1.4 V Jamie L. Cohen,† David J. Volpe,† Daron A. Westly,‡ Alexander Pechenik,§ and Hećtor D. Abrunã*,†,§ Department of Chemistry and Chemical Biology, Cornell University, Ithaca, New York 14853, Cornell Nanoscale Facility, Cornell University, Ithaca, New York 14853, and -92-6, LLC, Ithaca, New York 14850 Received August 19, 2004. In Final Form: January 28, 2005 A dual electrolyte H2/O2 fuel cell system employing a planar microfluidic membraneless fuel cell has been investigated and compared to single electrolyte H2/O2 systems under analogous conditions. The fuel is H2 dissolved in 0.1 M KOH (pH 13), and the oxidant is O2 dissolved in 0.1 M H2SO4 (pH 0.9), comprising a system with a calculated thermodynamic potential of 1.943 V (when 1 M H2 and O2 concentrations are assumed). This value is well above the calculated thermodynamic maximum of 1.229 V for an acid, or alkaline, single electrolyte H2/O2 fuel cell. Experimentally, open-circuit potentials in excess of 1.4 V have been achieved with the dual electrolyte system. This is a 500 mV increase in the open circuit potentials observed for single electrolyte H2/O2 systems also studied. The dual electrolyte fuel cell system shows power generation of 0.6 mW/cm2 from a single device, which is nearly 0.25 mW/cm2 greater than the values obtained for single electrolyte H2/O2 fuel cell systems studied. Microchannels of varying dimensions have been employed to study both the single and dual electrolyte H2/O2 systems. Channel thickness variation and the flow rate dependences of power generation are also addressed.

1. Introduction Recently, there has been great emphasis on the development of micro-fuel-cell technologies, requiring the ability to scale down macro-fuel-cell systems while retaining their power generation capabilities.1 Some of the greatest challenges involved have been the incorporation of a polyelectrolyte membrane (PEM) into a micro fuel cell and its long-term stability, especially under high current density operation.2 The PEM is a complex polymer membrane, often intimately associated with the anode and cathode catalysts, creating a complex system that must be decreased in size without diminishing its effectiveness. The PEM also has intrinsic problems, such as drying out (particularly at elevated operating temperatures), tearing, deterioration, and inefficient prevention of fuel crossover.3,4 Decreasing the membrane thickness exacerbates these problems. Therefore, recent work has been directed at eliminating the traditional PEM in micro-fuel-cell designs.5,6-12 Bio * Corresponding author: tel, +1-607-255-4720; fax, +1-607-2559864; e-mail address, [email protected]. † Department of Chemistry and Chemical Biology, Cornell University. ‡ Cornell Nanoscale Facility, Cornell University. § -92-6, LLC. (1) Maynard, H. L.; Meyers, J. P. J. Vac. Sci. Technol., B 2002, 20, 1287-1297. (2) Wainright, J. S.; Savinell, R. F.; Liu, C. C.; Litt, M. Electrochim. Acta 2003, 48, 2869-2877. (3) Savadogo, O. J. Power Sources 2004, 127, 135-161. (4) Larminie, J.; Dicks, A. Fuel Cells Explained; John Wiley & Sons: Ltd.: West Sussex, England, 2000. (5) Mitrovski, S. M.; Elliott, L. C. C.; Nuzzo, R. G. Langmuir 2004, 20, 6974-6976. (6) Choban, E. R.; Markoski, L. J.; Wieckowski, A.; Kenis, P. J. A. J. Power Sources 2004, 128, 54-60. (7) Ferrigno, R.; Stroock, A. D.; Clark, T. D.; Mayer, M.; Whitesides, G. M. J. Am. Chem. Soc. 2002, 124, 12930-12931. (8) Katz, E.; Willner, I.; Kotlyar, A. B. J. Electroanal. Chem. 1999, 479, 64-68.

fuel cells in this category use enzymes in order to achieve the selectivity needed to prevent the effects of fuel crossover from the anode to the cathode.8,10-12 While selectivity is attained, power generation is low and the problems inherent to biological systems, such as proper buffering and complicated immobilization techniques, require close monitoring, Nevertheless these bio fuel cells are being developed for low power applications. There has also been recent work reported on membraneless micro fuel cells based on laminar flow.6,7,9 The interface between the fuel and oxidant acts as a diffusive membrane, allowing ionic transport between the fuel and oxidant streams. Fuel crossover is essentially eliminated because the solutions mix only by diffusion at their interface, and fuel and oxidant are constantly flowing out of the cell at short time scales due to high flow rates employed. Choban and coworkers have demonstrated the feasibility of this concept using a Y-shaped microchannel design in which the fuel and oxidant are flowed side by side.6 In addition, we recently reported on planar microfluidic membraneless fuel cell design that employs the characteristics of laminar flow of the fuel and oxidant streams between two parallelplate electrodes.9 The planar electrodes allow for increased power generation due to the large contact area between the fuel and oxidant and the electrode surfaces. This design obviates the need for a PEM and the attendant problems associated with it, as well as allowing the dimensions of the micro fuel cell to be decreased with little complication. Membraneless micro-fuel-cell studies have typically focused on formic acid fuel cell systems, or vanadium redox chemistry, as a proof of concept.6,7,9 The power densities (9) Cohen, J. L.; Westly, D. A.; Pechenick, A.; Abrunã, H. D. J. Power Sources 2005, 139, 96-105. (10) Mano, N.; Mao, F.; Heller, A. J. Am. Chem. Soc. 2002, 124, 1296212963. (11) Mano, N.; Mao, F.; Heller, A. J. Am. Chem. Soc. 2003, 125, 65886594. (12) Mano, N.; Mao, F.; Shin, W.; Chen, T.; Heller, A. Chem. Commun. 2003, 518-519.

10.1021/la0479307 CCC: $30.25 © 2005 American Chemical Society Published on Web 03/16/2005

H2/O2 Fuel Cell System

Langmuir, Vol. 21, No. 8, 2005 3545

Figure 1. Thermodynamic considerations of H2/O2 systems. The acid employed is 0.1 M H2SO4 (pH 0.9) and the base is 0.1 M KOH (pH 13). Potentials vs standard hydrogen electrode (SHE).

reported for formic acid systems, as well as the power generation from any single micro-fuel-cell device reported to date, is lower than that required for the applications in which micro fuel cells would be of great value, such as cell phones and other small portable devices. Another fuel that has been studied extensively is pure hydrogen.4,13-15 Because it can be oxidized at very low overpotentials on platinum catalyst surfaces, this fuel can readily be employed as a model system to explore other aspects of membraneless micro fuel cells, such as geometry and flow rate. Likewise, hydrogen has a much higher energy conversion efficiency than most other fuels, thus increasing the power generation from the micro-fuel-cell device. In this report we employ the planar micro-fuel-cell platform, previously reported,9 to study three H2/O2 fuel cell systems. The first two systems are single electrolyte, acid or alkaline, systems that have been studied extensively.1,4,13,14 These systems have been shown to generate power efficiently and are attractive because they generate only H2O as a byproduct of the fuel cell reaction. The third system is a dual electrolyte H2/O2 system. This system uses H2 as a fuel dissolved in 0.1 M KOH and O2 as an oxidant dissolved in 0.1 M H2SO4. The main benefit that accrues from this system is the increased opencircuit potential (OCP) over the single electrolyte systems and the concomitant increase in power. This membraneless design eliminates the possibility of the acid and base streams homogenizing, and it allows for anion and cation conduction across the solution interface. The large pH difference between the two solutions, and the large ion concentration gradients, establishes a relatively large liquid junction potential between the fuel and oxidant, the role of which is discussed. The behavior of this dual electrolyte system with microchannels of varying dimensions, flow rates, and microchannel thicknesses is discussed and compared to the single electrolyte H2/O2 systems. 2. Thermodynamic Aspects Figure 1 shows the thermodynamic potentials expected for acidic and alkaline single electrolyte H2/O2 systems at room temperature (assuming unit activity/fugacity of dissolved/gaseous species, respectively).16 The acidic one (13) Acres, G. J. K. J. Power Sources 2001, 100, 60-66. (14) Gu¨lzow, E. J. Power Sources 1996, 61, 99-104. (15) Haile, S. M. Acta Mater. 2003, 51, 5981-6000.

is the most extensively researched H2/O2 system because it is employed in PEM fuel cells.3,4,13,15 The alkaline electrolyte system has also been studied, but to a lesser degree due to common problems associated with the generation of insoluble products (often referred to as carbonation) during the oxidation of the fuel, resulting in deterioration of the PEM, as well as electrode, performance.4,13,14 The maximum thermodynamically attainable voltage from these systems is 1.229 V when calculated using 0.1 M H2SO4 (pH 0.9) for the acid, 0.1 M KOH (pH 13) for the base, and assumed concentrations of 1.0 M for H2 and O2. Experimentally, the OCP for a H2/O2 fuel cell is ca. 0.8-0.9 V. This deviation is primarily due to the large overpotential associated with the reduction of O2 (despite the somewhat enhanced kinetics in alkaline media), but the decrease can also be attributed to other resistances in the fuel cell itself, including solution resistance and electrical contact resistance. Because a PEM is carefully engineered to enhance the mobility of specific ions,3,4,15,17 and macro fuel cells are typically designed to recycle the electrolyte such that a pH gradient cannot be maintained, the third type of H2/ O2 fuel cell described in Figure 1 has never, to our knowledge, been investigated. By use of the concentrations and assumptions previously noted, a dual electrolyte fuel cell with H2 dissolved in an aqueous alkaline solution and O2 dissolved in an aqueous acid solution has a thermodynamically calculated OCP of 1.943 V. This is 714 mV greater than that calculated for acid, or alkaline, single electrolyte systems. This unique system can be readily studied with a membraneless micro fuel cell because the diffusive interface between the fuel and oxidant is not specific to the type of ion(s) traversing the interface. It is also a flow cell, which mitigates the problem of fuel and oxidant mixing, which would result in neutralization of the base and acid. Even with the O2 overpotential, experimental OCPs can be expected to surpass the thermodynamic maxima of previously studied hydrogen fuel cells. Another advantage is that the aqueous waste products of this system can form a neutral solution, thus (16) Bard, A. J.; Faulkner, L. R. Electrochemical Methods; Fundamentals and Applications, 2nd ed.; John Wiley & Sons: New York, 2001. (17) Costamagna, P.; Srinivasan, S. J. Power Sources 2001, 102, 242252.

3546

Langmuir, Vol. 21, No. 8, 2005

Cohen et al.

Table 1. Acid Single Electrolyte H2/O2 Micro-Fuel Cell

a

channel widtha

open circuit potential (V)

total area (cm2)

power (µW)

power density (µW/cm2)

rel flow velocity (4.4 cm/s)

1 mm wide 1 mm wide, 250 µm thick 3 mm wide 5 mm wide,b 250 µm thick five-microchannel array

0.880 0.890 0.910 0.850 0.875

0.5 0.5 1.5 2.5 2.5

110 90 270 650 570

220 180 180 260 228

1.0 1.5 0.8 1.2 0.8

Channel thickness was 380 µm unless otherwise noted. b Si electrodes were employed in place of Kapton electrodes.

limiting the generation of strongly acidic or strongly basic hazardous waste. 3. Experimental Section Aqueous solutions of 0.1 M KOH (Fisher Scientific, Fair Lawn, NJ) or 0.1 M H2SO4 (J.T. Baker-Ultrapure Reagent, Phillipsburg, NJ), saturated for 30 min with H2 (Ultrapure, Airgas, Inc., Radnor, PA) prior to introduction into the fuel cell, were used as fuels. Aqueous solutions of 0.1 M KOH or 0.1 M H2SO4, saturated for 30 min with O2 (Airgas) prior to introduction into the fuel cell, were used as the oxidant. Millipore water (18 MΩ cm, Millipore Milli-Q) was used to make the aqueous acidic and alkaline solutions. All data were obtained using the fuel cell platform previously described. The cell design, fabrication, and instrumentation have been described in detail in previous work.9 Figure 2 shows a side

Figure 2. Side view of the planar microfluidic membraneless fuel cell. view of the device employed. Briefly, this platform consists of a silicon microchannel, 380 or 250 µm thick. These microchannels were fabricated with a “tapered flow boundary” less than 100 microns thick to aid in the establishment of laminar flow of the fuel and oxidant streams prior to bringing them into contact. Basic photolithographic techniques were employed in order to fabricate the Si microchannels which were 5 cm long and had widths of 1, 3, or 5 mm. Arrays of five microchannels, each channel 1 mm wide and 5 cm long, arranged in parallel and fed fuel and oxidant simultaneously, were fabricated as well. These microchannels were then placed between two flexible 300FN Kapton (Dupont, Wilmington, DE) electrodes, which acted as the anode and cathode. These electrodes consisted of 50 nm of Ta and 50 nm of Pt evaporated onto the 300FN Kapton surface, thus creating a large-area planar electrode surface. The microchannel, anode, and cathode were then clamped between two pieces of Plexiglas with eight bolts in order to apply an even pressure across the entire system, thus forming a watertight cell. Electrical contact to the electrodes was made at their ends, which protruded from the device, and copper foil was employed to decrease the contact resistance. The device was connected in series to a variable load resistor (HeathKit) and digital multimeter (Keithley, Cleveland, OH) in order to measure the voltage across the cell when a load was applied. Fuel and oxidant were pumped using a dual syringe pump (KD Scientific, Holliston, MA) with two syringes (Becton Dikinson lewer-lock 60 cm3) affixed with polyethylene tubing (2 mm o.d.) in order to integrate the pumping system to the fuel cell.

4. Results and Discussion 4.1. Acid and Alkaline Single Electrolyte H2/O2 Systems. As mentioned above, traditional H2/O2 fuel cells have both the fuel and oxidant dissolved entirely in acidic or entirely in alkaline aqueous electrolyte solutions. Figure 3 shows a current density vs potential (i-V) curve for a 1 mm wide, 380 µm thick microchannel fuel cell. The fuel and oxidant were aqueous solutions of 0.1 M H2SO4 saturated with H2 or O2, respectively, for 30 min. Table 1 presents data obtained for a variety of microchannel widths. From the shape of the curve (initial plateau), it was determined that the H2/O2 system was mass transport limited. The power generation was thus flow-rate dependent. That is to say that in the limit of fast mass transport down the length of the microchannel, the system will be kinetically limited. In addition, such higher flow rates require higher pumping rates, which may decrease the integrity of the water-tight cell, also limiting the power output of the device. Relative flow velocities used to obtain the power results reported are given in Table 1. Voltage losses were less than 400 mV, which can be mainly attributed to the O2 reduction overpotential. Using H2 as a fuel gives rise to large power enhancements over formic acid, which was the test fuel system reported previously using this planar fuel cell platform.9 A variety of microchannel dimensions were employed, and power generation was expected to scale linearly with increasing microchannel width. Table 1 shows that this indeed was the case. A single device with a 5 mm wide channel produced 0.65 mW of power, while a single 1 mm wide channel produced 220 µW/cm2. Figure 4 and Table 2 show the results obtained for an alkaline electrolyte H2/O2 fuel cell system. The fuel and oxidant were aqueous solutions of 0.1 M KOH saturated

Figure 3. i-V curve for a single 380 µm deep Si microchannel fuel cell. Kapton electrodes with 50 nm of Ta and 50 nm of Pt were used. Fuel was a solution of 0.1 M H2SO4 saturated with H2 and oxidant was a solution of 0.1 M H2SO4 saturated with O2. Flow rate was 1.0 mL/min.


Langmuir, Vol. 21, No. 8, 2005 3547

Figure 4. i-V curve for a single 380 µm deep, 1 mm wide, Si channel. Kapton electrodes of 50 nm of Ta and 500 nm of Pt were used. Fuel was a solution of 0.1 M KOH saturated with H2, and the oxidant was a solution of 0.1 M KOH saturated with O2. The flow rate was 1.0 mL/min.

Figure 5. i-V curve for a 1 mm wide, 380 µm deep, Si channel. Kapton electrodes of 50 nm of Ta and 50 nm of Pt were used. Fuel was 0.1 M KOH saturated with H2, and the oxidant was 0.1 M H2SO4 saturated with O2. The flow rates were 0.5 mL/ min (squares) and 2.0 mL/min (circles).

with H2 or O2, respectively, for 30 min prior to introduction into the fuel cell. The versatile platform employed allows interrogation of an alkaline system by simply changing the fuel and oxidant injected. There is no membrane to limit ion mobility, nor is the system plagued by the typical problems found in alkaline systems, such as buildup of insoluble carbonate, as the products are expelled from the system in less than 1 s at the flow rates typically employed. The ability to conveniently use alkaline electrolyte systems may reopen areas of alkaline fuel cell research that have previously been discounted due to the aforementioned problems. Open-circuit potentials in excess of 900 mV were obtained and power generation from a single device was nearly 1 mW. The alkaline system produced more power than the acid electrolyte H2/O2 system and had higher OCPs. Figure 4 shows the typical mass-transport-limited curve shape including a welldefined plateau, as well as greater initial current density at zero load than observed for the acid system. These data support the contention that there is a kinetic enhancement of the H2 oxidation and O2 reduction in an alkaline environment, with the latter likely providing the larger enhancement. 4.2. The Dual Electrolyte H2/O2 Fuel Cell System. The dual electrolyte fuel cell used 0.1 M KOH, saturated with H2 as the fuel, and 0.1 M H2SO4, saturated with O2 as the oxidant. Figure 5 shows data obtained for a 1 mm wide, 380 µm thick microchannel at two different flow rates. The dual electrolyte micro-fuel-cell data obtained using microchannels of varying widths are presented in Table 3. The OCPs were consistently in excess of 1.35 V with power generation more than twice that for single electrolyte H2/O2 fuel cell systems. Table 3 also demonstrates the importance of flow rate on power generation.

Figure 6. Power results for the data presented in Figure 5.

The shape of the profile at 0.5 mL/min (average velocity of 4.4 cm/s) looked much like that of a single electrolyte H2/O2 system demonstrating mass-transport-limited behavior. The power density generated at 0.5 mL/min was indeed analogous to the corresponding power density for an alkaline electrolyte fuel cell system. As the flow rate was increased to 2.0 mL/min (average velocity of 17.5 cm/ s), the initial current density was approximately the same as that for the 0.5 mL/min flow rate, but then it jumped to more than twice the initial value. This “jump” led to much larger power densities than could be obtained in previous H2/O2 systems, as shown in Figure 6. In fact, the power was quadrupled when the flow rate of the solutions was increased from 0.5 to 2.0 mL/min for both fuel and

Table 2. Alkaline Electrolyte H2/O2 Micro Fuel Cell

a

channel widtha


total area (cm2)

power (µW)



1 mm wide 1 mm wide, 250 µm thick 3 mm wide 5 mm wide,b 250 µm thick five-microchannel array

0.905 0.930 0.930 0.905 0.940

0.5 0.5 1.5 2.5 2.5

145 143 350 920 800

290 286 233 368 320

1.0 1.5 0.7 1.2 0.8

Channel thickness was 380 µm unless otherwise noted. b Si electrodes were employed in place of Kapton electrodes.

3548


Cohen et al.

Figure 7. Power results for a five-microchannel (each 1 mm wide, 250 µm thick) array. Kapton electrodes of 50 nm of Ta and 50 nm of Pt were used. Fuel was 0.1 M KOH saturated with H2, and oxidant was 0.1 M H2SO4 saturated with O2. The flow rate was 4.0 mL/min. Table 3. Average Power Results for the Alkaline/Acid Dual Electrolyte H2/O2 Micro Fuel Cell

channel widtha 1 mm wide

1 mm wide, 250 µm thick 3 mm wide 5 mm wide,b 250 µm thick five-microchannel array a


total area (cm2)

power (µW)



1.41 1.44 1.44 1.45 1.40 1.38 1.45 1.44

0.5 0.5 0.5 0.5 0.5 1.5 2.5 2.5

119 200 251 375 480 627 1500 1200

237 400 501 750 960 418 600 480

1.0 2.0 3.0 4.0 6.0 2.0 2.4 1.6

Channel thickness was 380 mm unless otherwise noted. b Si electrodes were employed in place of Kapton electrodes.

oxidant streams. The power density obtained from a single 1 mm wide channel was 750 µW/cm2. Table 3 also shows results for a variety of microchannels, and Figure 7 shows the power results for a five-microchannel array, which generated 1.2 mW. While the power densities for these microchannels were higher than those for the acid and alkaline single electrolyte H2/O2 systems, they did not scale linearly with electrode area. This was likely due to the fact that the 1 mm wide fuel cell was much further refined relative to the other microchannels. With additional refinements to the wider channels and the five-microchannel array, analogous results would likely be obtained and power outputs would likely scale with electrode area. The OCPs listed in Table 3 average 1.43 V. While this is more than 500 mV greater than those reported with the single electrolyte systems, it is noted that these OCPs reflect a loss of over 500 mV when compared to the thermodynamic expectations depicted in Figure 1. The thermodynamics of this dual electrolyte system were maximized for a H2/O2 fuel cell, but kinetics of this specific configuration are not optimal. The rate of oxidation of hydrogen in alkaline media is retarded by over an order

of magnitude compared to that in acid electrolyte.18,19 The overpotential of oxygen reduction in acid electrolyte is larger than that in alkaline electrolyte because of the deleterious effects of (bi)sulfate anion surface poisoning of the polycrystalline Pt (it is noted that while OH- can also adsorb to the Pt surface, the (bi)sulfate adsorption has been reported to have a greater deleterious effect of the reaction kinetics).20-23 Both of these factors contribute significantly to the potential loss in this system. Another deleterious effect is the liquid junction potential generated at the interface of these two solutions. It was calculated that a liquid junction potential on the order of 50 mV was generated.18,19 The liquid junction potential, slow kinetics of the H2 oxidation and O2 reduction, gas concentrations (18) Lagger, G.; Jensen, H.; Josserand, J.; Girault, H. H. J. Electroanal. Chem. 2003, 545, 1-6. (19) Josserand, J.; Lagger, G.; Jensen, H.; Ferrigno, R.; Girault, H. H. J. Electroanal. Chem. 2003, 546, 1-13. (20) Anderson, A. B. Electrochim. Acta 2002, 47, 3759-3763. (21) Markovic, N. M.; Ross, P. N. In Interfacial Electrochemistry: Theory, Experiment, and Applications; Wieckowski, A., Ed.; Marcel: New York, 1999; pp 821-841. (22) Markovic, N. M.; Gasteiger, H. A.; Philip N. Ross, J. J. Phys. Chem. 1995, 99, 3411-3415. (23) Wang, J. X.; Markovic, N. M.; Adzic, R. R. J. Phys. Chem. B 2004, 108, 4127-4133.


Figure 8. i-V curve for single 380 µm, 1 mm wide, Si channel. Kapton electrodes of 50 nm of Ta and 500 nm of Pt were used. Fuel was 0.1 M H2SO4 saturated with H2, and oxidant was 0.1 M KOH saturated with O2. The flow rate was 1.5 mL/min.

lower than those assumed when thermodynamic calculations were carried out, and solution resistances all contribute to the losses in potential for this system. Nonetheless, the important fact is that OCPs of up to 1.5 V consistently were achieved. The dual electrolyte system can also be run with an acidic anode stream and an alkaline cathode stream. Figure 1 shows that the thermodynamic potential of this system should be 0.515 V. While the system was not expected to have nearly the power generation of the previous dual electrolyte system discussed, characterization of its behavior was expected to provide additional insight into the behavior of dual electrolyte fuel systems. Figure 8 shows the results obtained for this type of system using a 1 mm wide, 380 µm thick microchannel. The curve shape for this dual electrolyte system was analogous to the single electrolyte H2/O2 system, which showed mass transport limitations. The anomalous shape exhibited in the i-V curves observed with an alkaline anode stream and acidic cathode stream was not present in this configuration. It is also noted that the system in Figure 8 had an OCP of 0.513 V. It was mentioned that the previous dual electrolyte system suffered from slow kinetics, which, in turn, diminished the OCP of the cell, even though the thermodynamics of the system were more favorable. In this system, the opposite scenario exists. The thermodynamics of the system are such that low OCPs will always be garnered from this dual electrolyte system, but the kinetics are optimized. The H2 oxidation reaction is reportedly much faster in acidic electrolyte and the oxygen reduction reaction kinetics are enhanced in alkaline media, in part due to the absence of Pt surface poisoning from the (bi)sulfate anion mentioned above.24-26 As mentioned previously, there was a liquid junction potential associated with the dual electrolyte system that was deleterious to the OCP produced by the system. The noteworthy aspect of this liquid junction is that, because it was not electrochemically generated, it need not inhibit power generation. The liquid junction potential was (24) Schmidt, T. J.; Stamenkovic, V.; Ross, P. N.; Markovic, N. M. Phys. Chem. Chem. Phys. 2003, 5, 400-406. (25) Markovic, N. M.; Gasteiger, H. A.; Philip N. Ross, J. J. Phys. Chem. 1996, 100, 6715-6721. (26) Gasteiger, H. A.; Markovic, N. M.; Philip N. Ross, J. J. Phys. Chem. 1995, 99, 8290-8301.

Langmuir, Vol. 21, No. 8, 2005 3549

Figure 9. Diagram of the liquid junction potential setup at the fuel/oxidant interface: (a) the liquid junction potential between an alkaline H2 anode and acidic O2 cathode; (b) the liquid junction potential between an acidic H2 anode and alkaline O2 cathode.

established due to the pH gradient within the channel at the interface of the two solutions, as well as along the microchannel length as the acid and base diffusively mixed. Figure 9a shows the typical liquid junction for the alkaline anode stream and acidic cathode stream dual electrolyte system described initially. On the basis of ion mobilities, a net positive charge was maintained in the anode solution stream, while the cathode stream retained a net negative charge. The junction potential thus engendered was determined to diminish the OCP, as well as power generation, of the dual electrolyte fuel cell. As described, the electrolyte configurations were reversed, and a liquid junction was established as depicted in Figure 9b. In this case, a net negative charge was expected to persist in the anode stream, facilitating oxidation at the anode, while a net positive charge near the cathode would enhance reduction at the cathode. Thus, the liquid junction potential in this case could be expected to increase the power of the cell. The enhanced kinetics, coupled with the favorable liquid junction potential, allows one to obtain OCPs at, or near, the thermodynamic maximum for this configuration of the dual electrolyte system. 4.3. Variations in Behavior due to Microchannel Thickness. Microchannels with a thickness of 250 µm were fabricated in order to determine if variations in the channel thickness caused variations in the power densities, open-circuit potentials, and current densities obtained for a particular fuel cell system. As mentioned previously, when 250 µm (as opposed to the 350 µm) thick microchannels were employed in the formic acid fuel system, there was no significant change in power generation. Tables 1 and 2 also show that the 250 µm microchannel performed analogously to the 380 µm thick microchannel in the single electrolyte H2/O2 system. Figure 10 shows i-V curves for a 380 µm microchannel and a 250 µm microchannel. The thinner microchannel exhibited a typical curve shape for a single electrolyte H2/O2 system that was mass transport limited. Note that the current density to which the 380 µm microchannel system increased was also the current density at the edge of the plateau of the 250 µm microchannel. This behavior suggests that the atypical curve feature can be noted as an initial “dip” in current density at the beginning of the experiment and that the current density actually recovered to the current density seen at the peak maximum in the i-V curve. Figure 11 shows the power density as a function of potential for these two systems. These data demonstrate that the irregular current density vs potential profile in the 380 µm microchannel system contributes significantly to the power output of the device.

3550


Figure 10. i-V curves for a single 380 µm (circles) and 250 µm (squares) deep, 1 mm wide, Si channels. Kapton electrodes of 50 nm of Ta and 500 nm of Pt were used. Fuel was 0.1 M KOH saturated with H2, and the oxidant was 0.1 M H2SO4 saturated with O2. Flow rate was 2.0 mL/min.

Figure 11. Power results for data presented in Figure 10.

5. Conclusions The versatility of a membraneless fuel cell has been demonstrated using three H2/O2 systems. Single electrolyte H2/O2 systems, using acid or base, were employed and mass-transport-controlled behavior was observed. Open circuit potentials between 0.850 and 0.940 V were obtained. These open-circuit values are comparable to those obtained in some of the most efficient conventional (macro) fuel cells currently being studied. It was also demonstrated that by using a variety of microchannels one can tailor the power generation of an individual device. Power generation of 650 µW in acid electrolyte and 920 µW in alkaline electrolyte was achieved. A new dual electrolyte H2/O2 fuel cell system that generated OCPs greater than 1.4 V was demonstrated.

Cohen et al.

The system utilized the negative oxidation potential of H2 dissolved in an alkaline electrolyte and the positive reduction potential of O2 when dissolved in an acidic solution. Significant power increases, compared to the acid and alkaline electrolyte systems alone, were obtained, with 1.5 mW generated from a single device. The OCPs observed were consistently more than 500 mV greater than those typically observed in single electrolyte fuel cells. Flow rate and channel thickness were determined to be factors in the power output of the devices, as well as the i-V curve shape. The establishment of a liquid junction potential was also determined and the magnitude of this potential was estimated to be on the order of 50 mV. This value, in conjunction with the kinetic effects of the electrolyte in which the H2 and O2 were dissolved, significantly affected the OCPs of the dual electrolyte systems. A system in which the alkaline electrolyte stream flowed at the anode and the acidic electrolyte stream flowed at the cathode had a liquid junction potential and slow kinetics that were deleterious to the OCP. In the reverse dual electrolyte system, the liquid junction potential and optimized kinetics contributed favorably to the system’s performance. Several points can be concluded from the dual electrolyte studies above: (a) this system established a liquid junction potential detrimental to its power generation; (b) decreasing the microchannel thickness eliminated the anomalous curve shape; (c) the anomalous curve shape was not observed at slow flow rates. While it seems that charge reorganization, as well as pH gradients along the length of the channel, were the main contributors to the unusual curve shape of this system, further interrogation is needed in order to clarify the exact cause of the initial “dip” in the current density and its subsequent increase to a current density maximum at more positive potentials. Future work will also focus on the role of geometric factors, such as channel length, in the dual electrolyte system, the ability to control, and further quantify, the liquid junction potential by varying the electrolyte concentrations, as well as the investigation of new fuels using this fuel cell platform. Finally, as with many flow fuel cells, it is recognized that fuel utilization is poor, and further work is being carried out in order to improve upon the fuel consumption of this device. The dual electrolyte system may be one of the many new fuel systems that are viable because of the unique advantages that this fuel cell design affords. Acknowledgment. This research was generously funded by ARO (DAAD19-03-C-0100) a Phase I STTR contract to Cornell and -92-6, LLC, and NSF (ACT0346377). This work was performed in part at the Cornell NanoScale Facility (a member of the National Nanotechnology Infrastructure Network) which is supported by the National Science Foundation under Grant ECS-0335765, its users, Cornell University, and Industrial Affiliates. Special thanks to Professor H. White at the University of Utah for useful discussions. LA0479307

O2 Planar Membraneless Microchannel Fuel

Recommend Documents