Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Dynamics of direct hydrocarbon polymer electrolyte membrane fuel cells
(USC Thesis Other)
Dynamics of direct hydrocarbon polymer electrolyte membrane fuel cells
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
DYNAMICS OF DIRECT HYDROCARBON POLYMER ELECTROLYTE MEMBRANE FUEL CELLS BY EUGENE H. KONG A DISSERTATION PRESENTED TO THE FACULTY OF THE USC GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS OF THE DEGREE DOCTOR OF PHILOSOPHY (MECHANICAL ENGINEERING) AUGUST 2019 2 Acknowledgments I would like to express my sincere gratitude to my advisor Professor Paul D. Ronney for his excellent tutelage, continuous support, as well as, friendship during my entire PhD program. Under his supervision, I learned how to perform a successful research and become the scientist I am today. I would also like to thank Professor G. K. Surya Prakash for giving me the privilege work in his lab and interact with high caliber scientists in his group. I am humbled, grateful, and indebted to both Professor Paul D. Ronney and Professor G. K. Surya Prakash for entrusting me with this project and making my dissertation possible. I would like to express a special acknowledgement to Professor Sri R. Narayan who has been invaluable in my research by introducing me to the field of electrochemistry and batteries. He has always been encouraging and giving me insights and valuable comments on the electrocatalysis aspects in my research. I am also grateful for thoughtful suggestions from my committee members: Professor Fokion N. Egolfopoulos, Professor Mitul Luhar and Professor Satwindar Sadhal. I would also like to acknowledge Donald R. Wiggins and Mike from the machine shop for their incredible machinery knowledge and always being accommodating of my requirements to make my experimental setup possible. Colleagues at Loker Hydrocarbon Institute have also made significant contributions to this research project. First I would like to thank Dr. Marc Iuliucci and Dr. Dean E. Glass for training me since the first day from membrane electrode fabrication to testing the fuel cell system. I would also like to thank Dr. Bo Yang, Vicente Galvan, Amanda Baxter, Sayan Kar, Sahar Roshandel, and Adam Ung for intellectual conversations and thoughtful inputs into my project. I would also like to thank my colleagues at Combustion Physics Laboratory, Dr. Ashkan Davani, Si Shen, 3 Jakrapop Wongwiwat, Harris Zhou, Brandie Rhodes, Yang Shi, Winry Patharapong, for stimulating discussions and all the fun we had together at combustion conferences. Most importantly, I would like express my gratitude to all my colleagues for their unpretentious friendships and encouragements when research became frustrating. Finally, I would like to convey my deep appreciation to my father, mother, and sister. This dissertation would not have been possible without their warm love, continued patience, and endless support. My parents’ selfless support from my childhood to now has made this work possible and is the most valuable gift in my life. I would like to dedicate this dissertation to my greatest heroes – Mom & Dad. 4 Table of Contents List of Figures 8 List of Tables 11 Abstract 12 Chapter 1 – Introduction and Literature Review 13 1.1. Why Hydrocarbon? 13 1.2. Motivation and Applications 15 1.3. Prior Studies of Propane Electro-oxidation 17 1.4. Objectives 20 Chapter 2 – Experimental Methodology 22 2.1. Fuel Cell Apparatus 22 2.2. Membrane Pretreatment and Electrode Fabrication Process 24 2.3. Experimental Apparatus and Procedures 25 Chapter 3 – Results and Discussion 28 3.1. Dynamics of Hydrocarbon PEMFCs 28 3.1.1. Fuel Purity Grade 28 3.1.2. Research Grade Propane & Unsaturated Hydrocarbons 29 3.1.3. Effect of Current Scan Rates 32 3.1.4. Current Dynamics of Hydrocarbon PEMFCs 37 3.1.5. Dynamics of Hydrogen and Dimethyl Ether PEMFCs 42 3.1.6. CO Mitigation and CO Bleeding 45 3.1.7. Effluent Analysis 49 5 3.1.8. Fuel Cell Extinguishment Model 59 3.2. Testing Parameters 65 3.2.1. Effect of Temperature 65 3.2.2. Effect of Membrane Thickness 68 3.2.3. Effect of Catalyst Loadings 70 3.2.4. Effect of Shear Stress 71 3.2.5. Effect of Fuel Types 76 3.2.6. Aluminum Bipolar Plates 80 3.3 Characterization Results 84 3.3.1 X-Ray Photoelectron (XPS), X-Ray Diffraction (XRD), and Scanning Electron Microscope (SEM) 84 Chapter 4 – Future Work 91 4.1. In Situ Spectroscopy Method 91 4.2. Platinum Treatment – Hydrazine 93 4.3. Catalyst Synthesis 100 Chapter 5 – Conclusion 101 References 103 6 List of Figures Figure 1.1. Maximum burner temperatures (center of spiral counter flow heat exchanger) at the extinction limits. Figure 1.2. Power density of direct PEMFCs with different catalysts. Propane flow rate = 1.2 L/min. Oxygen flow rate = 1 L/min. Cell temperature = 80˚C. Membrane: Nafion 117 [24]. Figure 2.1. Schematic drawing of a single fuel cell apparatus 1) current collector plates 2) graphite flow fields 3) gaskets to prevent gas or liquid leakage 4) carbon paper for gas diffusion layer to enhance diffusion of reactants 5) platinum black for electro-oxidation of propane 6) PEM for proton conductivity [27]. Figure 2.2. A schematic drawing of the fuel cell setup. Fuel and oxygen in bubbled through humidifiers at 90˚C controlled by the temperature controller and the data was measured using Fuel Cell Testing System connected to a computer. Figure 3.1 Dynamics of direct hydrocarbon PEMFCs at 80˚C using research-grade propane. Effect of bleeding 2540 ppm (0.254%) of unsaturated hydrocarbons at 36 mA/cm 2 (without the addition, no power is produced). Figure 3.2. Effect of bleeding different concentrations of ethylene at 36 mA/cm 2 (for 0 ppm, the cell was first “ignited” with ethylene then the ethylene flow was stopped). Note logarithmic time scale. Figure 3.3. Polarization curves of direct propane PEMFCs at different scan rates at 80˚C. The flow rate was 1.2 L/min and 0.8 L/min for propane and oxygen, respectively. Figure 3.4. Polarization curves of direct hydrogen PEMFCs at two different scan rates at 80˚C. The flow rate was 1.2 L/min and 0.8 L/min for hydrogen and oxygen, respectively. The same experimental setup as that of hydrocarbon PEMFC. Figure 3.5. Polarization curves of direct dimethyl ether PEMFCs at two different scan rates at 80˚C. The flow rate was 1.2 L/min and 0.8 L/min for dimethyl ether and oxygen, respectively. The same experimental setup as that of hydrocarbon PEMFC. Figure 3.6. Galvanostatic mode of the direct propane PEMFCs at various constant currents for 1000 seconds operating at 80˚C. Figure 3.7. Galvanostatic mode and load-interrupt mode of direct propane PEMFCs at low current density. The flow rate was 1.2 L/min and 0.8 L/min for propane and oxygen, respectively. The operating temperature was held at 80˚C. The current was applied for 20 seconds and shut off for 5 seconds. 7 Figure 3.8. Galvanostatic mode and load-interrupt mode of a 25cm 2 direct propane PEMFCs at medium current density. The flow rate was 1.2 L/min and 0.8 L/min for propane and oxygen, respectively. The operating temperature was held at 80˚C. The current was applied for 20 seconds and shut off for 5 seconds. Figure 3.9. Galvanostatic mode and load-interrupt mode of a 25cm 2 direct propane PEMFCs at high current density. The flow rate was 1.2 L/min and 0.8 L/min for propane and oxygen, respectively. The operating temperature was held at 80˚C. The current was applied for 20 seconds and shut off for 5 seconds. Figure 3.10. Polarization curves of direct propane PEMFCs with different operating schemes at 80˚C. The flow rate was 1.2 L/min and 0.8 L/min for propane and oxygen, respectively. The current was applied for 20 seconds and shut off for 5 seconds for load-interrupt mode. Figure 3.11. Polarization curves of direct DME PEMFCs with various operating schemes at 80˚C. The flow rate was 1.2 L/min and 0.8 L/min for dimethyl ether and oxygen, respectively. The scan rate was .667 mA/cm 2 s. Figure 3.12. Polarization curves of direct hydrogen PEMFCs with different operating schemes at 80˚C. The flow rate was 1.2 L/min and 0.8 L/min for hydrogen and oxygen, respectively. The scan rate was 4 mA/cm 2 s. Figure 3.13. Effect of oxygen bleeding on the current density. The pressure is P = 1 atm, temperature T = 80˚C and CO concentration of 100 ppm. Figure 3.14. Polarization curves of PEMFCs: propane, propane with addition of carbon monoxide (≈14.3%), CP grade carbon monoxide at 0.133 mA/cm 2 s at 80˚C. Figure 3.15. Constant current held at 36 mA/cm 2 for propane, propane with addition of carbon monoxide (≈14.3%), and pure CO at 80˚C. Figure 3.16. GC analysis of anode downstream at various current. Plot of carbon conversion ratio (ratio of fuel complete oxidation to carbon dioxide) in primary axis vs. current and total charge consumed during testing in secondary axis vs. current. Figure 3.17. Plot of total charge consumed vs. carbon conversion ratio. Figure 3.18. IR result of distilled water before the experiment and distilled water after the experiment. Figure 3.19. The direct propane PEMFCs extinguishment model based on mathematical derivation with constants shown in the figure (K 1 =2.4 & K 2 = 2). (The 𝐾 " 𝐶 $% & term leads to self-accelerating of the rate of formation of inactive sites which is required to obtain a prediction of extinction; see text). 8 Figure 3.20. The direct propane PEMFC extinguishment models based on mathematical derivation with constants K 1 =2.4, K 2 = 2, a = 2.5, K 3 = 6600 / (x+737) with x = 0. Figure 3.21. The direct propane PEMFC extinguishment models based on mathematical derivation with constants K 1 =2.4, K 2 = 2, a = 2.5, K 3 = 6600 / (x+737) with x = 1040. Figure 3.22. The direct propane PEMFC extinguishment models based on mathematical derivation with constants K 1 =2.4, K 2 = 2, a = 2.5, K 3 = 6600 / (x+737) with x = 1540. Figure 3.23. The direct propane PEMFC extinguishment models based on mathematical derivation with constants K 1 =2.4, K 2 = 2, a = 2.5, K 3 = 6600 / (x+737) with x = 2000. Figure 3.24. Polarization curves of direct propane PEMFC at different operating temperatures. The flow rate was 1.2 L/min and 0.8 L/min for propane and oxygen, respectively. The scan rate was .133 mA/cm 2 s. Figure 3.25. Arrhenius plot based on the experimental result of temperature effect. Figure 3.26. Polarization curves of direct propane fuel cell with various membrane thicknesses at 80˚C. The flow rate was 1.2 L/min and 0.8 L/min for propane and oxygen, respectively. The scan rate was .133 mA/cm 2 s. Figure 3.27. Polarization curves of direct propane PEMFCs with low and high catalyst loadings at 80˚C. The flow rate was 1.2 L/min and 0.8 L/min for propane and oxygen, respectively. The scan rate was .133 mA/cm 2 s. Figure 3.28. The power density curve of the same fuel cell: one with decreased flow channel volume and one without. Both tests were run on a hydrazine treated propane PEMFCs run at a constant current of 36 mA/cm 2 . Figure 3.29. Galvanostatic mode of direct propane PEMFCs at 24 mA/cm 2 at three different heights without hydrazine treatment. Figure 3.30. Galvanostatic mode of direct propane PEMFCs at 36 mA/cm 2 at three different fuel flow rate at the normal height without hydrazine treatment. The working temperature was at 80˚C. Figure 3.31. Polarization curves of direct propane n-butane, and isobutane PEMFCS scanning at 0.133 mA/cm 2 s and operating at 80˚C. Figure 3.32. Galvanostatic mode of direct propane, n-butane, and isobutene PEMFCs at 36 mA/cm 2 and operating at 80˚C. Figure 3.33. Polarization curve of direct ethylene PEMFC scanning at 0.133 mA/cm 2 s. The operating temperature was held constant at 80˚C and the same experimental setup and testing parameters (flow rate, catalyst loadings…etc.) were employed. 9 Figure 3.34. Galvanostatic mode at various currents for direct ethylene PEMFC at 80˚C. The operating temperature was held constant at 80˚C and the same experimental setup and testing parameters (flow rate, catalyst loadings…etc.) were employed. Figure 3.35. The bipolar plate for PEMFC application with active area 5 cm 2 and pin post type flow field was created in SolidWorks. . Figure 3.36. The complete aluminum fuel cell apparatus after being machined in the USC machine shop. Figure 3.37. The aluminum bipolar plates of the fuel cell after a couple hours of testing. Figure 3.38. A closer view of the aluminum plate after testing for a couple of hours. Figure 3.39. XPS (Kratos Axis Ultra) result of new platinum catalyst with Mono Aluminum X- ray working at x -10 Torr, power setting: 6 mA and 10 kW. Figure 3.40. XPS (Kratos Axis Ultra) result of platinum catalyst after electrocatalysis with Mono Aluminum X-ray working at x -10 Torr, power setting: 6 mA and 10 kW. Figure 3.41. XRD (Rigaku Ultima IV) result of new platinum catalyst and platinum after electrocatalysis using 𝜅−𝛼 X-ray working at 40 kV, 44 mA, and 1.76 kW. Figure 3.42. SEM (JSM-6610LV low vacuum) images of new platinum catalyst and platinum catalyst after electrocatalysis. Upper Left: New platinum at low magnification. Upper Right: Platinum after electrocatalysis at low magnification. Middle Left: New platinum at medium magnification. Middle Right: Platinum after electrocatalysis at medium magnification. Bottom Left: New platinum at high magnification. Bottom Right: Platinum after electrocatalysis at high magnification. Figure 4.1. Schematic drawing of in-situ time resolved FTIR. Figure 4.2. Scanning electron microscopy (SEM) pictures of platinum foil before (left) and after heating in NH 3 (right) – air mixture for an hour. The bare Pt foil was burned with propane–air mixtures with ≈ 5% of the propane replaced by ammonia for one hour. (field of view 6µm x 8µm). Figure 4.3. XRD patterns for the untreated Pt black and treated Pt catalyst. Figure 4.4. Cyclic voltammograms of PtBl and Pt(N 2 H 4 ) catalysts. Measurements were performed in 0.5 M H 2 SO 4 solution saturated by Ar. Scan rate: 20 mV/s. Figure 4.5. Cyclic voltammograms of propane oxidation on PtBl and Pt(N 2 H 4 ) catalysts. Measurements were performed in 0.5 M H 2 SO 4 solution under propane flow at 90˚C. Scan rate: 20 mV/s. 10 Figure 4.6. Polarization curves of direct propane PEMFC with various anode catalysts at 80°C. Propane pressure and flow rate: 1 atm and 1.2 L/min, oxygen pressure and flow rate: 1 atm and 0.5 L/min. Both streams were humidified at 90˚C [29]. Figure 4.7. The platinum catalyst with hydrazine treatment process recommended by Hong et al. As shown, serious agglomeration of the catalyst. 11 List of Tables Table 3.1. Total number of electrons reacted per propane molecule at various currents. Table 3.2. The efficiency of fuel utilization at various currents. Table 3.3. Platinum binding energy for 4f7/2 values. Table 4.1. Average particle size and SA of untreated Pt black and treated Pt catalysts. 12 Abstract Hydrocarbon fuels contain ≈ 50 times more energy per unit mass than commercially available batteries, thus harvesting only 10% of this energy content could provide an improved power source for portable electronic devices. With this motivation, the feasibility of using polymer electrolyte membrane fuel cells with hydrocarbon fuels, operating at low temperatures (< 100˚C), was explored. As for membrane electrode assembly, Nafion® N-117 was used for electrolyte and platinum black powder was used as catalyst. With extremely pure (>99.99%) propane no power is produced, however, with the addition of trace quantities of unsaturated hydrocarbons, power production ensues and continues even after the unsaturated hydrocarbon is discontinued. Furthermore, the current history has a significant influence on the cell performance. In particular, at higher current densities (> 24 mA/cm 2 ) the power output gradually decreases then rapidly “extinguishes,” however, by periodically shutting off the current for time intervals, the average power density increased significantly. A heuristic model considering the relative rates of conversion of active anode catalyst sites to inactive sites and vice versa was developed to interpret this “extinguishment” behavior. This model was in good qualitative agreement with experimental data; possible physical mechanisms for this heuristic model are discussed. 13 Chapter 1 – Introduction and Literature Review 1.1. Why Hydrocarbons? It is well known that most fuels contain far more energy per unit mass than batteries [1], hence in recent years many attempts have been made to create portable electrical power sources using fuels as the energy storage mediums [2]. Such sources could replace batteries in some applications, with the advantages of far greater power and/or lifetime, instant recharge ability, no “memory effect” and having no toxic material disposal issues. Polymer electrolyte membrane fuel cells (PEMFCs), electrochemical devices that convert chemical enthalpy directly into electricity, are a natural choice for fuel-based portable power devices because of their low operating temperatures, rapid startup, simple construction, efficiency, quietness, and clean exhaust compared to other forms of electrical power generators [3-4]. Currently, hydrogen is the most common fuel for PEMFCs because of low operating temperature and high power density (≈ 0.7 W/cm 2 ) [5]. However, hydrogen is highly reactive, extremely flammable and difficult to store at high ratios of energy/mass or energy/volume ratios when the mass or volume of the storage medium is included [6]. Thus, the direct methanol PEMFCs have been introduced to avoid the difficulties of fuel storage and many improvements have been made over the past decade [7-9], but one main challenge for direct methanol PEMFCs is that methanol closely resembles water (which the membrane must allow to flow across) leading to fuel crossover losses from the anode to cathode. The toxicity of aldehydes formed during methanol oxidation is another concern. Furthermore, they are both (hydrogen & methanol) manufactured from natural gas using complex reactor systems that have large capital cost and have high cost and weight infrastructure for distribution and storage [10]. 14 The direct hydrocarbon PEMFCs could be the most practical type of fuel cells for portable electrical power systems because of the storability (inexpensive and easily available infrastructure), relatively low flammability hazard, and lack of toxicity of many hydrocarbons. Some portable power systems have been proposed using hydrocarbons or alcohols that are reformed into hydrogen and carbon monoxide (CO) and then fed into the fuel cell after eliminating CO, but the reformer system adds considerable weight, volume, complexity, and additional cost of at least 30% of the total cost of the fuel cell system [11-12]. Despite many advantages over other types of fuel cells, the direct hydrocarbon PEMFCs have been overlooked because hydrocarbon reaction rates on fuel cell anodes at low temperature are one or more orders of magnitude slower than those of methanol and hydrogen [13]. 15 1.2. Motivation and Applications The current work on the direct hydrocarbon PEMFCs started from prior work on mesoscale catalytic combustion reactors, which has shown that the ammonia-treated platinum can sustain hydrocarbon and dimethyl ether (DME) reaction at low temperature (≈ 77˚C vs. ≈ 647˚C) as shown in figure 1.1. Figure 1.1. Maximum burner temperatures (center of spiral counter flow heat exchanger) at the extinction limits [14]. Combustion can be considered as a shorted fuel cell in the sense that the reactants and product are the same, only the intermediate steps are different. If hydrocarbons can sustain combustion at 75˚C, will it possible to make direct hydrocarbon PEMFCs working at 75˚C? Typically, hydrocarbon fuel cells require much higher temperatures to have sufficient reaction rates, hence the use of solid oxide fuel cells has been widely accepted operating in the range of 400˚C – 800˚C. To maintain elevated temperature environment, spiral counter flow heat exchanger has been introduced by placing the solid oxide fuel cell in the center to minimize heat loss and it 16 has shown to be a very effective implementation [15]. But there are large heat losses, thermal stresses, limits materials choices when fabricating spiral counter flow heat exchanger due to high temperature. It would be beneficial to use hydrocarbons in a fuel cell at less than 100˚C so that atmospheric pressure PEMFCs could be also implemented in spiral counter flow heat exchanger. Some advantages operating at a low temperature include; less heat loss and a wide variety of materials for fabrication – particularly, plastics with low thermal conductivity to minimize heat losses, and cheap fabrication. As mentioned before, direct hydrocarbon PEMFCs have advantages over conventional primary (nonrechargeable) batteries or secondary batteries (long times to recharge) because hydrocarbons have 50 to 100 times higher energy densities and they can be considered as instantaneously rechargeable batteries. Thus, even at a relatively low operating efficiency of 10%, they will still provide far more energy at lower costs than those of commercial batteries [16]. After a successful development of the direct hydrocarbon PEMFCs, this technology would have a huge impact on the national defense. For example, the U.S. Army sends some soldiers into combat with as much as 20 pounds of batteries to power their electronic systems; being able to reduce their burden to 2 pounds would represent a major increase in the soldier’s efficacy and survivability in hostile situations. Another application of the direct hydrocarbon PEMFCs is for the transportation systems. Using direct hydrocarbon PEMFCs for vehicles could potentially have much better energy efficient than combustion engines and would be essentially zero-emissions – practically no CO or soot particles are emitted. 17 1.3. Prior Studies of Propane Electro-oxidation The very first few works on the direct anodic oxidation of hydrocarbon for fuel cell applications have been introduced in the 1960s in hopes to obtain fuel cells that can carry useful current densities and have only water and carbon dioxide as byproducts [17-19]. Fuels such as, not only limited to hydrocarbons; propane, methane, ethane, butane, pentane, hexane, octane, decane, hexadecane, propylene, ethylene, and carbon monoxide were studied and different electrolytes such as sulfonated phenolformaldehyde acid and aqueous cesium salt were studied. Niedrach demonstrated propane fuel cell using sulfonated phenolformaldehyde acid electrolyte and achieved maximum power density of ~ 1.5 mW/cm 2 with 18 mg/cm 2 of loading at 85˚C [17]. After, Grubb et al. demonstrated propane fuel cell operating at a range of 150˚C to 200˚C. It was concluded that phosphoric acid (H 3 PO 4 ) was the most effective electrolyte and maximum current density of 50 mA/cm 2 at 0.5V cell potential at 200˚C was achieved [18]. Finally, Cairns demonstrated high performance propane fuel cell with aqueous cesium salt as electrolyte at catalyst loading of 50 mg/cm 2 at 150˚C to achieve maximum power density of 80 mW/cm 2 [19]. It has been generally concluded that platinum black as for the catalyst was favored over palladium [17] and the propane showed the highest oxidation rate among hydrocarbons [19]. It was also concluded that all hydrocarbons showed similar performances and they were all by far inferior compared to that of hydrogen, and propane showed slightly better performance than that of propylene [17]. Further investigation on the direct propane PEMFCs has been reported by Savagado and Rodriguez using solid electrolytes instead of aqueous electrolytes [20-24]. Instead of using strong aqueous acid or solutions as electrolytes, they studied H 2 SO 4 doped polybenzimidazole (PBI) membranes [20]. They reported that with Pt-CrO 3 /C catalyst on anode, a maximum power density of 46 mW/cm 2 was achieved and with Pt-Ru/C, a maximum power density of 42 mW/cm 2 was 18 observed at 95˚C [22]. Later in their work, they also studied Dupont TM Nafion® N-117 membrane as electrolyte and with 20% PtOx/C anode catalyst, maximum power density 14.5 mW/cm 2 was observed and with 40% Pt/C anode catalyst, maximum power density of 6.48 mW/cm 2 was reported at 80˚C as shown in figure 1.2. Li et al. also reported viability of direct propane PEMFCs with Dupont TM Nafion® N-117 as electrolyte and he reported maximum power densities of 1.47 mW/cm 2 for propane, 1.33 mW/cm 2 for butane, and 0.69 mW/cm 2 for ethane at 90˚C [25]. Figure 1.2. Power density of direct PEMFCs with different catalysts. Propane flow rate = 1.2 L/min. Oxygen flow rate = 1 L/min. Cell temperature = 80˚C. Membrane: Nafion 117 [24]. The work by Savagado and Li et al. successfully reported working direct propane PEMFCs at a low temperature. However, previous investigators did not explore the time dependence of the fuel cell, which has a significant effect on the performance of the fuel cell, e.g. faster scan rates yield higher power densities. It is assumed that the time dependence is caused by polymerization 19 of some species on the catalyst surface leading to sudden extinction behavior. Another important parameter is the importance of unsaturated hydrocarbons (UH) in the fuel stream. It has been observed that very small amount of UH are needed to start the fuel cell and also very small amount of UH has a significant effect on the performance of the fuel cell. These unusual behaviors and mitigation methods from extinguishment will be discussed in this paper. 20 1.4. Objectives More efforts on the direct hydrocarbon PEMFCs have been made in the Combustion Physics Laboratory (CPL) to address aforementioned behaviors not addressed by previous investigators. Even at a low reaction rate of hydrocarbons compared to those of hydrogen and methanol, it has been found that the direct hydrocarbon PEMFCs are possible, but exhibit very unusual behaviors. At relatively high currents, the power output of hydrocarbon fuel cells is not stable and abruptly decreases to zero over a time scale of minutes and the time for the cell to abruptly extinguish decreases with increase in current, as opposed to hydrogen PEMFCs where the cell exponentially decays to zero at relatively high currents. It is assumed that this is due to poisoning of catalyst by some intermediate species during the oxidation of the fuel. This poisoning effect can be mitigated in two ways: bleeding a small amount of unsaturated hydrocarbons (UH) or operating the fuel cell in a load-interrupt mode. Bleeding a small amount of UH not only avoids extinguishment, but it is also a crucial to get the fuel cell started. It has been discovered that the direct hydrocarbon PEMFCs are not possible with high purity grade (research grade, >99.99% purity) propane. First a small amount of UH is necessary to get the fuel cell stated, but can be removed after. Another method to avoid the extinguishment is to employ an unconventional mode of operation. The load-interrupt mode enables the fuel cell to avoid extinguishment and increases the average power density (1.5 times higher) compared to that of galvanostatic mode (constant current), where the fuel cell is tested for 1000 seconds in both operating modes for each current. The current working hypothesis is that that there are some species getting adsorbed onto platinum surface and blocking propane molecules from reaching the catalyst, but this can be mitigated by adding UH to the fuel stream or turning off the current. The objective of this study is to maximize the power output from the direct 21 hydrocarbon PEMFCs and understand the unusual behavior when operating at low temperatures. More research and study of different parameters are necessary to elucidate mechanism(s) of the extinguishment of the direct propane PEMFCs behavior. The first part of this paper will discuss about the dynamics of the hydrocarbon PEMFCs and attempt to understand its unusual behavior by creating a heuristic model of the experimental behavior. The second part of the paper will discuss the effects of testing parameters of the fuel cell such as temperature, catalyst loadings, membrane thickness, effect of shear stress, and effect of fuel types. The last part of this paper will discuss results obtained by characterization instruments such as X-ray photoelectron (XPS), X-ray diffraction (XRD), and scanning electron microscope (SEM) of the platinum catalyst before and after the electrocatalysis. 22 Chapter 2 – Experimental Methodology 2.1. Fuel Cell Apparatus The fuel cell apparatus used in this work is composed of six main parts: current collector plates, graphite bipolar plates, Teflon gaskets, gas diffusion layer, catalyst layer, and polymer electrolyte membrane (PEM). The current collector plates are connected to the external loads to produce electricity and they also provide good electrical contact with graphite bipolar plates. Graphite plates were used as oppose to aluminum or other metals because they assure good conductivity and high resistance to acid [26]. The Teflon gaskets (PolyTetraFluoroEthylene film made with Teflon® fluoropolymer, thickness 0.005 inches) were necessary to act as barriers for potential fuel leaks and also work as sealing agents for fuel cell assembly. For this study, Toray Carbon Paper 060 – TGP-H-060 (thickness of .19mm with 10% wet proof) was used as gas diffusion layers, which are conductive porous materials to enhance reactant gas diffusion. As for the catalyst, platinum black from Sigma-Aldrich fuel cell grade (surface area 25-34 m 2 /g) was used for the electro-catalyst for both anode and cathode, while Dupont TM Nafion® N-117 membrane (conductivity ≥ 10 -2 S/cm) was used for the polymer electrolyte membrane (otherwise noted). The purpose of PEM is to separate the reactant gases and to transfer proton from anode to cathode. The schematic drawing of the fuel cell apparatus is shown in figure 2.1. As shown in the figure, fabrication of gas diffusion layer, catalyst layer, and polymer electrolyte membrane is called the Membrane Electrode Assembly (MEA), which will be discussed more in the next section. 23 Figure 2.1. Schematic drawing of a single fuel cell apparatus 1) current collector plates 2) graphite flow fields 3) gaskets to prevent gas or liquid leakage 4) carbon paper for gas diffusion layer to enhance diffusion of reactants 5) platinum black for electro-oxidation of propane 6) PEM for proton conductivity [27]. In the order shown in figure 2.1, each components were assembled together and held in place with eight long bolts and nuts. In order to prevent electrons flowing from anode to cathode via metal bolts, all bolts are equipped with plastic sleeves on both ends. Heating pads were also attached at both current collector plates (not shown) to maintain desired operating temperature. The heating pad was purchased from McMaster-Carr (part #35475K163). 24 2.2. Membrane Pretreatment and Electrode Fabrication Process For MEA fabrication, Dupont TM Nafion® membrane was first pretreated by boiling the membrane in 3% aqueous hydrogen peroxide (H 2 O 2 ) solution for one hour to remove organic and inorganic impurities contained in the membrane, boil it in deionized water for one hour, boil it in 5% sulfuric acid (H 2 SO 4 ) for one hour to protonate the membrane, and then again boil in deionized water for one hour to remove any impurities and debris prior to fabrication. The purpose of this treatment is to exchange Na + with H + in the sulfonic acid chains for better proton conductivity as commercial Dupont TM Nafion® are purposely built in SO 3- Na + form for easy processing, storage and transportation. The platinum catalyst ink was prepared by first adding one-part platinum black powder into a vial, three-part deionized water, and one-part Nafion® perfluorinated resin solution (5 wt. % in lower aliphatic alcohols and water, contains 15-20% water) to create binding between the catalyst and the membrane. The ink was then sonicated for 480s to disperse the solution. After the pretreatment and ink preparation, the ink was then directly painted onto Toray Carbon Paper 060 using a paint brush. The desired loading on the carbon paper has been determined by weighing the carbon paper prior to painting and after the painting process. The carbon paper with catalyst was then hot pressed onto the Dupont TM Nafion® membrane (PHI Model No. 50R1818S-2HCS- L-Y 1 S7) at 500 lbs & 120˚C for 20 minutes to form the MEA (refer to figure 2.1 components 4, 5, and 6 for a schematic drawing). 25 2.3. Experimental Apparatus and Procedures A schematic drawing of the fuel cell testing setup is shown in figure 2.2. It consists of; (1) distilled water and a pump for the humidifiers (2) home-built humidifiers with heat pads and temperature controller to operate humidifiers at desired operating temperature (3) flow meters (Matheson FM 1050-602-E300) to control the flow rate of incoming fuel and oxygen (4) fuel cell apparatus as described in the previous section (figure 2.1). Prior to testing, the fuel cell was preconditioned with dry nitrogen fed into the humidification system at ~90˚C for one hour to humidify the MEA for better proton conductivity and to purge the fuel cell system and humidifiers. The resistance across the fuel cell using the 4 probe impedance was measured to be 12 mΩ. The temperature of the fuel cell was measured and recorded with a thermocouple. After the fuel cell reached the desired operating temperature, the nitrogen was then shut off and humidified fuel and oxygen flew into anode and cathode, respectively. The operating temperature of the fuel cell was held at 80˚C for all tests otherwise noted, and the flow rate of humidified fuel and oxygen were set at 1.2 L/min for anode and 0.8 L/min for cathode with no back pressure on anode or cathode. 26 Figure 2.2. A schematic drawing of the fuel cell setup. Fuel and oxygen in bubbled through humidifiers at 90˚C controlled by the temperature controller and the data was measured using Fuel Cell Testing System connected to a computer. The fuel cell performance was measured using a Fuel Cell Testing System 890B (Scribner Associates, INC) connected to a computer. This system was operated in the galvanostatic mode, that is, a fixed current is applied to the fuel cell and the corresponding voltage recorded. Three different operating modes were employed: (1) current scan, in which the current is increased at a certain rate (mainly used to show data for hydrogen and dimethyl ether) (2) galvanostatic mode where constant current is applied to the fuel cell for 1000 seconds and (3) load-interrupt (unconventional) where the current is applied for a period of time and shut off for a few seconds and repeat the cycle for 1000 seconds. The last of these modes is not typically employed in fuel 27 cell testing but was found to be an effective mode of operation for the direct propane PEMFCs, which will be discussed more detail in chapter 3. 28 Chapter 3 - Results and Discussion 3.1. Dynamics of Direct Hydrocarbon PEMFCs In this section of the paper, the dynamics of direct hydrocarbon PEMFCs will be explored. The dynamics discussed in this section is interesting because these effects are irrelevant in most type of PEMFCs, but they have a huge effect on hydrocarbon PEMFCs at low temperatures. As discussed, the performance of the fuel cell is time dependent and this is not studied or discussed by previous investigators. The time dependence might be possibly caused by polymerization of some species on the catalyst surface, but this effect can be mitigated by employing several methods. 3.1.1. Fuel Purity Grade Our very first work on the direct hydrocarbon PEMFCs started with using chemically pure (CP grade, >99% purity) propane. However, it has been notice that different CP grade propane tanks show different effects on the fuel cell. For example, one CP grade propane tank starts the fuel cell immediately (OCV ~ 0.85V), but other CP grade propane tanks might take at least one hour for OCV to slowly increase to ~0.85V. Nothing in the experimental setups, testing procedures, or testing parameters were changed prior to changing the fuel. It was assumed that the impurities in the propane might be the causing this effect. As some CP grade propane tank might be purer than those of others. In combustion, 1 % of impurity in the fuel won’t have any effect in the results as fuel will be completely oxidized, but in PEMFCs, it has been verified by many researchers and investigators that even a few hundreds parts per million (ppm) of CO can poison the platinum sites, thus higher purity grade propane was used. Although CO is not in the CP grade 29 propane tank, it was assumed that other impurities in the fuel might be posioning catalyst, thus retarding the activation of catalyst sites. Thus, extremely high purity grade was employed for all experiments. 3.1.2. Research Grade Propane & Unsaturated Hydrocarbons Extremely pure (research grade, >99.99%) propane fuel does not produce any power whatsoever; the PEMFC simply does not “start” or “ignite”, open circuit voltage (OCV) ≈ 0.05V, at temperatures up to 90˚C. According to PRAXAIR catalog, impurities in the CP grade propane include: methane (≈500 ppm), ethane (≈700 ppm), butane (≈4000 ppm) and propylene (≈600 ppm) [28]. Only impurity other than hydrocarbons was propylene, thus ppm amount of propylene was purposely bleed into the research grade propane fuel stream. 0 5 10 15 20 25 1 10 100 1000 Power Density (mW/cm 2 ) Time (Seconds) Acetylene Isobutylene Propylene Ethylene 30 Figure 3.1. Dynamics of direct hydrocarbon PEMFCs at 80˚C using research-grade propane. Effect of bleeding 2540 ppm (0.254%) of unsaturated hydrocarbons at 36 mA/cm 2 (without the addition, no power is produced). First, propylene (2540 ppm) was bleed into the fuel stream and it enabled the fuel cell to start (OCV ≈ 0.85V) and produce power for over 1000 seconds at a constant current of 36 mA/cm 2 . Other types of UH were also studied, and the most effective tested UH thus so far is ethylene as shown in figure 3.1. It is interesting to note that the fuel cell can still operate even after a small trace of UH is removed (only needed in the beginning to get the fuel cell “started” or “ignited”), but the performance of the fuel cell can be enhanced with continuous feed of UH in the fuel stream. It is speculated that small concentration of UH is needed to get the fuel cell started and enhances the performance for the follwing reasons: 1. It is easier to oxidize UHs compared to hydrocarbons and since UHs oxidation is an exothermic reaction, it “triggers” propane oxidation. 2. Some species might be adsorbed on the platinum surface and UHs are helping to remove those species from the platinum surface. 3. Ethylene might be reacting with oxygen (crossover from cathode and oxygen is fatal to hydrocarbon PEMFCs) to form ethanol or DME, which are much easier to oxidize than that of propane. 31 Figure 3.2. Effect of bleeding different concentrations of ethylene at 36 mA/cm 2 (for 0 ppm, the cell was first “ignited” with ethylene then the ethylene flow was stopped). Note logarithmic time scale. The performance of the fuel cell relative to the concentration of ethylene in the fuel stream was studied as shown in figure 3.2. At 36 mA/cm 2 , without any addition of UH (0 ppm, first started with UH and removed), the fuel cell is inoperative past ≈ 70 seconds, but with addition of only 2000 ppm (0.2%) of ethylene can be operative past 1000 seconds (increasing maximum current density). Furthermore, with small amount of unsaturated hydrocarbon, the fuel cell can start even at room temperature, though with low open circuit voltage (0.6V) and the maximum power density of 0.5 mW/cm 2 (not shown). Otherwise noted, for all experiments, ethylene (≈ 2000 ppm) was bleed into fuel stream for 5 seconds to get the fuel cell started and shut off before starting the experiment. 0 5 10 15 20 25 30 1 10 100 1000 Power Density (mW/cm 2 ) Time (Seconds) 1540 ppm 1040 ppm 0 ppm 690 ppm 470 ppm 2540ppm 2000 ppm 32 3.1.3. Effect of Current Scan Rates In the course of conducting these fuel cell tests, it was found that the scan rates have a significant effect on the performance. While the direct propane PEMFCs have been studied by other scholars, no discussion on the time dependence of the fuel cell’s performance is reported. Specifically, no information about the scan rate of their polarization curve was discussed either, which is an important testing parameter for the direct propane PEMFC. As shown in figure 3.3, at the slowest scan rate (0.002 mA/cm 2 s), the power density is about 8 mW/cm 2 but at a faster scan rate (1.33 mA/cm 2 s), the power density almost doubles to 15.4 mW/cm 2 . This phenomenon can be explained by intermediate species accumulating over time on the reaction sites, blocking propane molecules from reaching the catalyst sites or making reaction sites inactive by poisoning the catalyst, causing the performance to decrease over time. Furthermore, at a really quick scan rate of 36 mA/cm 2 s, the power density can even reach 50 mW/cm 2 , but extinguish very quickly as well. From this graph, it can be observed that scanning twice as fast does not necessary increase the power density by twice as well. At scanning rates of 0.53 mA/cm 2 s and 1.33 mA/cm 2 s, the power densities were 13.1 mW/cm 2 and 15.4 mW/cm 2 , respectively. By almost scanning twice as fast, there was ~20% increase on the maximum power density and maximum current density. Even though it might not be significant amount, it is not negligible. This kind of behavior is not observed in any other types of fuel cells and have not been reported before. 33 Figure 3.3. Polarization curves of direct propane PEMFCs at different scan rates at 80˚C. The flow rate was 1.2 L/min and 0.8 L/min for propane and oxygen, respectively. The scan rates of different fuels for PEMFCs show that scan rates do not affect the performance for hydrogen and dimethyl ether PEMFCs as shown in figure 3.4 & figure 3.5. Two different scan rates were tested for each fuel. It was observed that for both fuels the performances were identical to each other even when the scan rates were significantly different. Unlike propane, where changing the scan rate can have a significant effect, hydrogen and dimethyl ether are not affected by time. 0 10 20 30 40 50 60 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 20 40 60 80 100 120 140 160 180 200 Power Density (mW/cm 2 ) Voltage (V) Current Density (mA/cm 2 ) 0.002 mA/cm 2 s 0.53 mA/cm 2 s 1.33 mA/cm 2 s 4 mA/cm 2 s 6.66 mA/cm 2 s 36mA/cm 2 s 34 Figure 3.4. Polarization curves of direct hydrogen PEMFCs at two different scan rates at 80˚C. The flow rate was 1.2 L/min and 0.8 L/min for hydrogen and oxygen, respectively. The same experimental setup as that of hydrocarbon PEMFC. 0 50 100 150 200 250 300 350 400 450 0 0.2 0.4 0.6 0.8 1 1.2 0 200 400 600 800 1000 1200 1400 1600 1800 Power Density (mW/cm 2 ) Voltage (V) Current Density (mA/cm 2 ) 4 mA/(cm^2 s) 8 mA/(cm^2 s) -‐5 5 15 25 35 45 55 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 50 100 150 200 250 300 Power Density (mW/cm 2 ) Voltage (V) Current Density (mA/cm 2 ) 0.133 mA/(cm^2 s) 0.4 mA/(cm^2 s) 35 Figure 3.5. Polarization curves of direct dimethyl ether PEMFCs at two different scan rates at 80˚C. The flow rate was 1.2 L/min and 0.8 L/min for dimethyl ether and oxygen, respectively. The same experimental setup as that of hydrocarbon PEMFC. Polarization curve is a widely accepted method to evaluate the performance of electrochemical devices such as fuel cells and batteries. For hydrogen PEMFCs, the performance is independent of scan rates and relatively long duration between each data point has a minimum effect on the polarization curve. Contrary to most PEMFCs, direct hydrocarbon PEMFCs is heavily time dependent and the duration between each measurement is critical to the polarization curve. As a result, galvanostatic mode was studied. For current densities 4 mA/cm 2 , 16 mA/cm 2 , 24 mA/cm 2 , steady-state behavior existed as shown in figure 3.6. Above current density of about 24 mA/cm 2 , the power slowly decreased over time then after some period of time rather quick drop to zero power, a phenomenon that we have termed “extinguishment” by analogy with flames. In flames, a small drop in temperature can have a significant effect on the chemical kinetics, similar behavior is assumed for direct hydrocarbon PEMFCs to cause the abrupt extinguishment behavior. For relatively large current densities, the rate of decrease of power was faster and the time to extinguishment was shorter. 36 Figure 3.6. Galvanostatic mode of direct propane PEMFCs at various constant currents for 1000 seconds operating at 80˚C. As shown in figure 3.6 galvanostatic mode testing for 1000 seconds at various currents might be more suitable for direct hydrocarbon PEMFCs to mitigate the time dependence effect as much as possible. At 28 mA/cm 2 , the fuel cell extinguishes before 1000 seconds and shows abrupt extinguishment behavior as discussed. All fuel cells will extinguish at relatively high currents (approaching limiting current density) due to mass transport losses, but the behavior is rather slow exponential decay behavior as opposed to a abrupt extinguishment behavior. The mass transport losses happen near limiting current density due to decrease in reactant concentration at the surface of the electrodes as fuel is used. At the limiting current density, the concentration at the catalyst 0 5 10 15 20 25 1 10 100 1000 Power Density (mW/cm 2 ) Time (Seconds) 16 mA/cm 2 28 mA/cm 2 32 mA/cm 2 36 mA/cm 2 8 mA/cm 2 4 mA/cm 2 37 surface is practically zero, as the reactants are consumed as soon as they are supplied to the surface, which eventually causes the fuel cell to extinct. It is also important to note that the power density curves drop from high power density to lower power density in the beginning of each test. This is caused by overshooting of current by the fuel cell test system and requires a few seconds for the current stabilize to desired operating currents. 3.1.4 Current Dynamics of Hydrocarbon PEMFCs Since the performance of the fuel cell is dependent on time, the performance of scan current mode is not also consistent with the galvanostatic mode. The results are significantly different in power density and current density. Theoretically, the overall propane fuel reaction is described in equations shown below. C 3 H 8 + 6H 2 O → 3CO 2 + 20H + + 20e - (Anode) 5O 2 + 20H + + 20e - → 10H 2 O (Cathode) C 3 H 8 + 5O 2 + 6H 2 O → 3CO 2 + 10H 2 O (Overall) Although the exact intermediate reaction steps during oxidation of propane mechanism inside the fuel cell in not well known, it is predicted that there might be alternative steps leading to formation of different species or CO which get adsorbed on the platinum surfaces blocking propane molecules from reacting with the platinum catalyst. It is well understood that even a small trace amount of CO in the fuel cell can have a significant effect on the fuel cell performance [30]. One operating scheme to circumvent absorption of intermediate species on platinum surface is to employ unconventional mode of operation, where the fuel cell is operated for a certain 38 period of time and the current is shut off and once again apply the current (load-interrupt mode). Remarkably, it was observed that merely setting the current to zero for a few seconds “reset” the fuel cell in the sense that when the same current was applied again, the power density was restored to the value at the beginning of the test, as if the first test and subsequent extinction event had never occurred. With the load-interrupt mode, it has shown that the performance can significantly increase compared to that of the galvanostatic mode. One explanation to this improvement is that when the current is shutoff the platinum surface might be getting desorb with intermediate species that might be blocking the propane from reaching the platinum surface. As shown in figures (figure 3.7 & figure 3.8) below, the average power densities for 12 mA/cm 2 and 24 mA/cm 2 with the galvanostatic mode were 6.23 mW/cm 2 and 7.61 mW/cm 2 , respectively. However, with the load-interrupt operating scheme the average power densities were 6.04 mW/cm 2 and 9.36 mW/cm 2 for 12 mA/cm 2 and 24 mA/cm 2 , respectively. At a low current, the load-interrupt operating scheme was not beneficial but at the higher current the power density was almost 1.25 times higher. In figure 3.7, 3.8, & 3.9, the experiments were tested for 1000 seconds but only shows up to 500 seconds for better examination. 39 Figure 3.7. Galvanostatic mode and load-interrupt mode of direct propane PEMFCs at low current density. The flow rate was 1.2 L/min and 0.8 L/min for propane and oxygen, respectively. The operating temperature was held at 80˚C. The current was applied for 20 seconds and shut off for 5 seconds. 0 2 4 6 8 10 12 14 16 0 50 100 150 200 250 300 350 400 450 500 Power Density (mW/cm 2 ) Time (Seconds) Load-‐Interrupt: 12 mA cm^-‐2 Constant Current: 12 mA cm^-‐2 0 5 10 15 20 25 30 0 50 100 150 200 250 300 350 400 450 500 Power Density (mW/cm 2 ) TIme (Seconds) Load-‐Interrupt: 24 mA cm^-‐2 Constant Current: 24 mA cm^-‐2 40 Figure 3.8. Galvanostatic mode and load-interrupt mode of direct propane PEMFCs at medium current density. The flow rate was 1.2 L/min and 0.8 L/min for propane and oxygen, respectively. The operating temperature was held at 80˚C. The current was applied for 20 seconds and shut off for 5 seconds. It was observed that at any higher current above 24 mA/cm 2 the fuel cell extinguishes before 1000 seconds, but with load-interrupt mode, the current density can be further increased to 40 mA/cm 2 and still be able to provide power for more than 1000 seconds. Calculated from figure 3.9, the average power density was 11.6 mW/cm 2 , which is almost 1.5 times higher than the average power density obtained through the galvanostatic mode at 24 mA/cm 2 (7.61 mW/cm 2 ). Figure 3.9. Galvanostatic mode and load-interrupt mode of direct propane PEMFCs at high current density. The flow rate was 1.2 L/min and 0.8 L/min for propane and oxygen, respectively. The operating temperature was held at 80˚C. The current was applied for 20 seconds and shut off for 5 seconds. 0 5 10 15 20 25 30 35 40 0 50 100 150 200 250 300 350 400 450 500 Power Density (mW/cm 2 ) Time (Seconds) Load-‐Interrupt: 40 mA cm^-‐2 Constant Current: 40 mA cm^-‐2 41 With this concept, the load-interrupt operating scheme was operated at 0.995 mA/cm 2 , 6.53 mA/cm 2 , 9.38 mA/cm 2 , 11.9 mA/cm 2 , 13.4 mA/cm 2 , 17.1 mA/cm 2 , 18.9 mA/cm 2 , 22.9 mA/cm 2 , 26.3 mA/cm 2 , 29.4 mA/cm 2 , 30.8 mA/cm 2 , 37.2 mA/cm 2 , and 41.1 mA/cm 2 (average current density, power density and voltage plotted) to compare the performance of two different operating modes (load-interrupt and galvanostatic). Aforementioned, it has shown that the load-interrupt mode has a significant advantage over the galvanostatic mode as shown in figure 3.10. Figure 3.10. Polarization curves of direct propane PEMFCs with different operating schemes at 80˚C. The flow rate was 1.2 L/min and 0.8 L/min for propane and oxygen, respectively. The current was applied for 20 seconds and shut off for 5 seconds for load-interrupt mode. As shown in figure 3.10, each data point for the galvanostatic mode and load-interrupt mode was held for 1000 seconds at various currents and the average value was plotted on the graph. This strongly signifies that direct propane PEMFCs is time dependent and the performance 0 2 4 6 8 10 12 14 16 0 0.2 0.4 0.6 0.8 1 0 10 20 30 40 50 Power Density (mW/cm 2 ) Voltage (V) Current Density (mA/cm 2 ) Load-‐Interrupt Galvanostatic 42 of the fuel cell decreases over time with increase in current. The extinguishment of fuel cell can be avoided by utilizing the load-interrupt operating scheme where current flow is stopped for a few seconds and the power returns to the initial value to obtain maximum power density of 11.6 mW/cm 2 . With this unconventional mode of operation, the load-interrupt mode yields higher maximum power density than that of the galvanostatic mode. The current working hypotheses is that there may be occasional alternate reaction pathways leading to formation of other species or CO on the anode which eventually poisons the anode catalyst, but gets desorb once the current is turned off. 3.1.5. Dynamics of Hydrogen and Dimethyl Ether PEMFCs The dynamics of other fuels is investigated. It is well known that hydrogen PEMFCs are not time dependent as it will be shown in this section. In addition, the dynamics of dimethyl ether (CH 3 OCH 3 , DME) is also studied because DME can be another candidate for transportation fuel. Although DME is inferior to propane in terms of energy density (28.8 MJ/kg vs 46.4 MJ/kg) and cost, it does have its merits. DME is the simplest ether, colorless, chemically stable gas and liquid which can be stored in high density liquid phase at around 5 atm [5]. Additionally, DME is known to have relatively high density and less explosive compared to those of methanol and hydrogen [36, 37]. One of main advantages of DME over propane or other hydrocarbon fuels is that it lacks C-C bond which is likely to form other organic compounds besides CO 2 that might poison the catalyst surface. Furthermore, due to lack of C-C bonds, it is easier to completely oxidize at a low temperature and reform to hydrogen with PEMFCs. The result shows that DME is a good fuel for PEMFCs and works well at a low temperature as shown in figure 3.11. The maximum power density was 48.8 mW/cm 2 compared to 10 mW/cm 2 for propane PEMFCs. Similar to propane 43 PEMFCs experiments, for the galvanostatic mode, each data point was tested for 1000 seconds and for the load-interrupt mode, load was applied for 20 seconds and off for 5 seconds for 1000 seconds. Another major advantage of DME compared to propane and other hydrocarbons is that it is not time dependent like the propane PEMFCs. For hydrocarbon PEMFCs, it is suspected that load-interrupt mode is superior compared to that of galvanostatic mode because there are some species adsorbed on the platinum surface, however, for DME, load-interrupt operating mode is not beneficial and it exhibits same performance as the galvanostatic mode. From this result, it can be assumed that there is no species adsorbed onto the platinum surface for DME PEMFCs. Figure 3.11. Polarization curves of direct DME PEMFCs with various operating schemes at 80˚C. The flow rate was 1.2 L/min and 0.8 L/min for dimethyl ether and oxygen, respectively. The scan rate was .667 mA/cm 2 s. 0 10 20 30 40 50 60 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 50 100 150 200 250 300 350 Power Density (mW/cm 2 ) Voltage (V) Current Density (mA/cm 2 ) Load-‐Interrupt Scan Current Galvanostatic 44 It can be seen in figure 3.11 that at high constant currents (>160 mA/cm 2 ), galvanostatic mode and load interrupt mode do not exactly match the scan current graph. This is because due to reaction rate limit and mass transport loss, and the fuel cell will extinguish at relatively high currents. For fuel cells not depending on time will exhibit exponential decay behavior and both load-interrupt mode and galvanostatic mode, each data point was held for 1000 seconds as opposed to 5 seconds for scan current graph. Thus, at high currents for load-interrupt and galvanostatic graph, the average power density will be slightly lower. Figure 3.12. Polarization curves of direct hydrogen PEMFCs with different operating schemes at 80˚C. The flow rate was 1.2 L/min and 0.8 L/min for hydrogen and oxygen, respectively. The scan rate was 4 mA/cm 2 s. The same test was also performed for the hydrogen fuel cells as shown in figure 3.12. As for hydrogen, the galvanostatic mode was significantly showed better performance than that of 0 50 100 150 200 250 300 350 400 450 0 0.2 0.4 0.6 0.8 1 1.2 0 400 800 1200 1600 Power Density (mW/cm 2 ) Voltage (V) Current Density (mA/cm 2 ) Load-‐Interrupt ScanCurrent Galvanostatic 45 load-interrupt mode. This is because there is no benefit for hydrogen PEMFCs to shutoff the current as there are no adsorbed species on the platinum surface and operating with load-interrupt mode will only hurt the overall performance. The slight deviation of galvanostatic mode from the scan current mode at high currents is because each data point is held for 1000 seconds for galvanostatic mode and only 5 seconds for the scan current mode. As expected, the load-interrupt mode does not benefit the hydrogen fuel cell. Unlike propane, hydrogen does not have carbon atom which might poison the platinum and there are far less intermediate species and reactions to completely oxidize hydrogen. Similar to propane fuel cell experiments, for the galvanostatic mode, each data point was tested for 1000 seconds and for the load-interrupt mode, load was applied for 20 seconds and off for 5 seconds for 1000 seconds. 3.1.6. CO Mitigation and CO Bleeding It has been long recognized that PEMFCs are very sensitive to CO poisoning and even a small trace can have a significant effect on the fuel cell performance. Phosphoric acid fuel cell can circumvent this difficulty by operating at ~150˚C, where CO adsorption and contamination of the catalyst is greatly reduced [18]. However, Dupont TM Nafion® membranes are preferred to operate at 80˚C to assure good proton conductivity [31] and at temperature above 120˚C, Nafion membranes are known to go through glass transition phase. At a low temperature (< 100˚C), the bond between carbon monoxide and platinum catalyst is very strong and prevents fuel from reaching the catalyst sites, thus reducing the overall performance of the fuel cell and the adsorption of carbon monoxide occurs slowly over time [42]. For CO mitigation, besides from operating at high temperatures, Gottesfeld and Pafford first proposed oxygen bleeding into the fuel stream to mitigate CO poisoning and experimentally 46 proved that injecting 2-6% of oxygen into the fuel stream can have a significant effect on the fuel cell performance [44]. Since then, this method has been believed to be one of the most effective CO mitigation methods. Many scholars explored oxygen or air bleeding numerically and experimentally and resulted that small amount of oxygen in the fuel stream can significantly improve the overall performance [42]. The reaction kinetics of heterogeneous oxidation CO on platinum surface is proposed by a Langmuir-Hinshelwood mechanism [45, 46]. Oxygen Adsorption O 2 + 2Pt ↔ O 2 -Pt + Pt → 2(O-Pt) CO Adsorption CO + Pt ↔ CO-Pt CO Oxidation CO-Pt + O-Pt → CO 2 + 2Pt As a result, CO can be oxidized into CO 2 by oxygen and lessen the poisoning effect by removing the adsorbed CO and restore the electrode surface area. By implementing numerical solution software, FEMLAB, Zamel and Li showed that even 0.5% and 1% of oxygen bleeding into the fuel with CO concentration 100 ppm can have a significant impact on the fuel cell performance as shown in figure 3.13 [43]. 47 Figure 3.13. Effect of oxygen bleeding on the current density. The pressure is P = 1 atm, temperature T = 80˚C and CO concentration of 100 ppm [43]. These studies were all performed with hydrogen as fuel, but this effective CO mitigation method can be certainly implemented to direct hydrocarbon PEMFCs where CO poisoning might be a problem and believed to be the cause of the fuel cell extinction. By injecting ~ 0.5% of oxygen into the fuel stream, it suspected that the direct propane PEMFCs will not be time dependent and avoid fuel cell extinguishment. However, for hydrocarbon PEMFCs, it has been discovered that even a small amount of oxygen in the anode stream immediately renders the PEMFCs inoperative; the amount required is roughly equal to the electron-receiving capacity of the oxygen flow based on 100% coulombic efficiency of the anode for oxygen, meaning that, unsurprisingly, the anode has far more affinity for oxygen (if present) than hydrocarbon fuel. In addition, platinum/ruthenium catalyst was 48 employed on the anode to suppress CO poisoning, but there was no improvement in the performance of the fuel cell. To study the the effect of CO on the performance of direct hydrocarbon PEMFCs, significant amount of CO (≈14.3%) was purposely bleed into the fuel stream. It can be expected for the fuel cell to extinct quicker, but surprisingly similar to those of UH, it helps to avoid extinction as shown in figure 3.14 and 3.15. Figure 3.14. Polarization curves of PEMFCs: propane, propane with addition of carbon monoxide (≈14.3%), CP grade carbon monoxide at 0.133 mA/cm 2 s at 80˚C. 49 Figure 3.15. Constant current held at 36 mA/cm 2 for propane, propane with addition of carbon monoxide (≈14.3%), and pure CO at 80˚C. CO as an additive to hydrocarbon fuels has the same qualitative effect as UH, but UHs are effective in much smaller concentration and yield higher power densities. Instead CO can actually be used as fuel, albeit with much lower power densities (≈ 4 mW/cm 2 ) than those of hydrocarbons as shown in figure 3.14. These results suggest that CO is not the cause of the sudden extinguishment behavior of the fuel cell and there might be other species poisoning the catalyst, but UH and even CO is reacting with those species to prevent poisoning of the catalyst. 3.1.7 Effluent analysis The purpose of this work is to investigate electro-oxidation of propane molecules on a platinum surface to determine any other intermediate species that might be blocking propane 0 5 10 15 20 25 30 1 10 100 1000 Power Density (mW/cm 2 ) Time (Seconds) Propane Propane + CO CO 50 molecules from reaching the catalyst sites. A partial electrochemical reaction of propane has been studied experimentally and computationally by many scholars [38-42]. Propane Oxidation Mechanism [38, 39] Propane Adsorption C 3 H 8 (g) → C 3 H 8(ads) Dehydrogenation of Adsorbed Propane C 3 H 8(ads) → C 3 H 7(ads) + H + + e - Water Adsorption H 2 O(g) → H 2 O (ads) Water Dissociation H 2 O (ads) → OH (ads) + H + + e - A few intermediate species that might be formed after adsorption of propane and water on platinum surface might include, i.e. C-C Bond Cleavage C 3 H 8(abs) → CH 4 + C 2 H 4 Carboxylic Acid Formation CO (abs) + H 2 O → COOH (ads) + H + + e - Propanol Formation C 3 H 7(ads) + OH (ads) → C 3 H 7 OH (ads) 51 It is of interest to study whether hydrocarbon reaction intermediates remain on the catalyst surface as oppose to complete oxidization to CO 2 by series of dehydrogenation (removal of hydrogen) and hydroxylation (gain of oxygen) reactions. While numerous steps still remain unknown, mechanism proposed by Vafaeyan et al. & Sustersic et al. suggest that some intermediate species might include carboxylic acid and propanol during electro-oxidation of propane molecule. It has also been reported by Vafaeyan et al. that by using density functional theory, the formation of propanol is unlikely due to higher activation required to form propanol instead of the complete oxidation to CO 2 on a Ni (100) anode catalyst surface [38]. However, this study was conducted on a Ni (100) surface and the result might be different for platinum and on different surfaces on platinum catalyst. By integrating GC and IR, the exhaust components of the fuel cell (under loading operation and OCV) will be analyze to detect carboxylic acid, propanol, methane, ethylene and other intermediate species. Those data will be compared to that of blank propane to ensure that those species were not included in fuel stream prior to the fuel cell operation. The result of these tests will hopefully generate more information on electro-oxidation of propane and gain insights on intermediates species formed during this reaction process. With this information, each intermediate species can be “seeded” (ppm amount) in the fuel stream to investigate on which species cause fuel cell extinction. As discussed in the previous section, CO is not contributing to the extinction behavior of the fuel cell. In this study by analyzing species in both gas and liquid phase by employing GC and IR, we are hoping to discover other unexpected intermediate species formed during oxidation of propane. 52 To maximize the possibility of detecting the intermediate species, the residence time of the reaction was increased by closing the valve on the downstream and upstream of the fuel cell anode stream so that limited amount of fuel is fully oxidized until the extinction (no flow of fuel). Figure 3.16. GC analysis of anode downstream at various current. Plot of carbon conversion ratio (ratio of fuel complete oxidation to carbon dioxide) in primary axis vs. current and total charge consumed during testing in secondary axis vs. current. The GC analysis on anode exhaust has shown only carbon dioxide and unreacted propane as major species and as for minor species, a very low concentration of methane and ethane (< 0.1%) were detected. At a low current (0.1A), on average, 50% of the fuel is completely oxidized into carbon dioxide before the extinguishment as shown in figure 3.16. However, as current increases, it is interesting to note that there is a steep drop in the carbon conversion ratio (ratio of conversion from propane to carbon dioxide) until 0.3A. It can be suspected that, at lower currents, 0 50 100 150 200 250 300 350 0 0.1 0.2 0.3 0.4 0.5 0.6 0 0.2 0.4 0.6 0.8 1 Charge (C) Carbon Conversion Ratio Current (Amps) Average Carbon Conversion Ratio Average Charge Consumed 53 there is enough time for fuel cell to “recover” from poisoning of the catalyst, but as current increases, there is not enough time for fuel cell to “recover” from the poisoning effect. The carbon conversion ratio was calculated by the following: , -. / , - 0 1 2 =𝐶𝑃 𝑅𝑎𝑡𝑖𝑜 (3.1) ,; <&=>? ,; <&=>?@" =𝐶𝑎𝑟𝑏𝑜𝑛 𝐶𝑜𝑛𝑣𝑒𝑟𝑠𝑖𝑜𝑛 𝑅𝑎𝑡𝑖𝑜 (𝐶𝐶 𝑅𝑎𝑡𝑖𝑜) (3.2) 𝐶 , I / :𝐶𝑜𝑛𝑐𝑒𝑛𝑡𝑟𝑎𝑡𝑖𝑜𝑛 𝑜𝑓 𝐶𝑂 N 𝑓𝑟𝑜𝑚 𝐺𝐶 𝐶 , 0 Q 2 :𝐶𝑜𝑛𝑐𝑒𝑛𝑡𝑟𝑎𝑡𝑖𝑜𝑛 𝑜𝑓 𝐶 " 𝐻 S 𝑓𝑟𝑜𝑚 𝐺𝐶 The 3 was added at the denominator because for every propane molecule, 3 CO 2 molecules are reacted. In the figure 3.16, it is suggesting that about ~ 50% of the propane is completely oxidize into the CO 2 at 0.1 A, but this data is misleading (although it does not effect electrons reacted per propane molecule calculation as shown later in this section). The volume enclosed in the fuel cell apparatus at the flow field is ~ 3.8 × 10 Z[ 𝑚 " and the total volume (including gases in the pipe) is ~ 6.5 × 10 Z[ 𝑚 " . The ratio of between these two volume is 0.58 and the fuel conversion ratio at 0.1 A is 0.51. It is suspected that close to ~100% of the fuel is actually completely oxidized into CO 2 because it is highly unlikely that the propane in the tube outside of the fuel cell apparatus will travel through the tubes to reach the catalyst sites only by diffusion. So when those unreacted propane molecules in the pipes are subtracted from calculation, 100% carbon conversion ratio has to be assumed to match the moles of electrons produced at 0.1 A. 54 Figure 3.17. Plot of total charge consumed vs. carbon conversion ratio. There is also a similar trend to the total charge (current * time until extinction) vs. current graph because as shown in figure 3.17, it shows a linear relationship. As expected, with increase in the carbon conversion ratio, we have increase in the total charge produced. y = 600.63x -‐22.16 R² = 0.97578 0 50 100 150 200 250 300 350 0 0.1 0.2 0.3 0.4 0.5 0.6 CHarge (C) Carbon Conversion Ratio 55 Figure 3.18. IR result of distilled water before the experiment and distilled water after the experiment. Infrared spectroscopy (IR) was also used to possibly detect any intermediate species dissolved in water as shown in figure 3.18, but no species dissolved in water were detected. This is also the limitation of IR. Since the most of the sample is water, it is hard to detect small trace of intermediate species in the sample if it is present. Two methods can be used to circumvent this challenge: 1. Extract the organic compounds in the sample using organic solvents (diethyl ether or ethyl acetate) a. Add diethyl ether in the sample to extract organic compounds (like dissolves like). 30 50 70 90 110 130 0 1000 2000 3000 4000 5000 Transmittance (%) Wavenumber (1/cm) Distilled Water Distilled Water from Exhaust 56 b. Dry the water in the sample using magnesium sulfate or sodium sulfate (no water is allowed in GC-MS). c. Eject the sample in GC-MS and detect organic compounds in the sample. 2. Employ High-Performance Liquid Chromatography (HPLC) a. This instrument pumps the mixture in a solvent at high pressure through a column (filled with solid adsorbent material) and has the ability to separate and identify compounds that are present in any sample in trace concentration as low as parts per trillion. The first method was employed using both diethyl ether and ethyl acetate but no intermediate species were detected in the GC-MS. The second method was not possible because the current HPLC machine is not currently working. Unfortunately, no other intermediate species were detected both in gas phase and liquid phase to explain the sudden extinguishment behavior of the fuel cell. This might suggest that the surface of the catalyst might be poisoned by hydrocarbon itself, polymerizing on the surface and blocking further reactions from happening. With enclosed volume (downstream & upstream closed), finite fuel, and knowing the carbon conversion ration, electrons reacted per propane molecule can be calculated with the Faraday’s constant. 𝑃𝑉 =𝑛𝑅𝑇 (3.3) The ideal gas law is used to calculate the mole of propane inside the anode side of the fuel cell and multiply by the carbon conversion ratio to calculate the total mole of propane reacted. To calculate the mole of electrons: 𝐶𝑢𝑟𝑟𝑒𝑛𝑡 × 𝑇𝑖𝑚𝑒 =𝑇𝑜𝑡𝑎𝑙 𝐶ℎ𝑎𝑟𝑔𝑒 (3.4) 57 Once, the total charge is calculated, it can be divided the Faraday’s constant to calculate the mole of electrons reacted. Current (Amps) Number of Electrons per Propane Molecule 0.1 27.2 0.18 25.1 0.3 20.1 0.5 19.9 0.7 20.4 0.9 23.1 Table 3.1. Total number of electrons reacted per propane molecule at various currents. Before this experiment, it was speculated that at most only two electrons will react from one propane molecule because the temperature is too low to completely oxidize hydrocarbons. However, results show that there are no intermediate species and propane molecules to completely oxidize to carbon dioxide. As shown in table 3.1, at high currents current (> 0.18A), about 20 electrons react per propane molecule, which is the maximum number of electrons that can react. This result is in a good accordance with the GC and IR result because there were no intermediate species detected and the number of electrons per propane molecule calculated close to theoretical number of electrons per propane molecule. However, at low currents (0.1 A & 0.18 A), the calculation shows that there are more than 20 electrons per propane molecule, which is not possible. Some explanations to explain these discrepancies might be due to the followings: 1. There is an internal small amount of current flowing in the fuel cell testing system itself (noise) contributing to the result. Even though the current is very small, propane PEMFCs is working with small current and over a long duration of testing, 58 the extra current adds up and contributing to extra electron count per propane molecule. 2. There might be fuel (propane) leakage somewhere in the fuel cell system and over a long duration of testing, a significant amount of fuel might have escaped from the system. Thus, the concentration of propane is less during GC testing. Current (Amps) Efficiency (%) 0.1 73 0.18 56.5 0.3 37.6 0.5 29.5 0.7 27.4 0.9 25.4 Table 3.2. The efficiency of fuel utilization at various currents. The efficiency of the fuel utilization was also calculated as shown in table 3.2. The heat of enthalpy of propane at 80˚C was calculated to be 2064 kJ/mol and the mol of propane can be calculated as described in the beginning of this section. By multiplying them together we can get the total enthalpy of propane that reacted in the fuel cell. Next, the amount of energy produced can be calculated by following: 𝑇𝑜𝑡𝑎𝑙 𝐸𝑙𝑒𝑐𝑡𝑟𝑖𝑐 𝑊𝑜𝑟𝑘 = 𝑉×𝐼 =>hi =? ij=>klm>%nhik= o 𝑑𝑡 (3.5) 𝐸𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑐𝑦= r?=&s tsiu=v>u w?vx r?=&s tk= n&syz ?{ ;v?y&ki <i&u=i| (3.6) For the total electric work, the power density was fitted with a polynomial fit and the equation was obtained from Microsoft Excel. 59 3.1.8. Fuel Cell Extinguishment Model In an attempt to understand and interpret this unusual result and behavior of the dynamics of hydrocarbon PEMFCs, a simple and heuristic model of the fuel cell was created. The most notable and unusual behaviors incorporated in the model include: 1. The abrupt extinguishment behavior of the fuel cell at relatively high currents as opposed to exponential decay behavior for hydrogen and DME PEMFCs. 2. A small concentration (ppm amount) of UH significantly improving the performance of the fuel cell. First it is assumed that there are only two types of sites on the fuel cell anode, namely active sites and inactive sites, whose concentrations (in moles per cm 2 of anode active area) are denoted as 𝐶 &% and 𝐶 >% , respectively. The sum of these two types of sites is constant, i.e. 𝐶 = =𝐶 &% +𝐶 >% (3.5) where 𝐶 = is the total concentration of sites on the anode. Next, it is assumed that the rate at which the concentration of active sites recovers from inactive sites (i.e., the source of active sites) is proportional to the number of inactive sites, i.e. |, ~, |= =𝐾 𝐶 >% (3.6) 𝐶 &%,%?mvui :𝑆𝑜𝑢𝑟𝑐𝑒 𝑜𝑓 𝑎𝑐𝑡𝑖𝑣𝑒 𝑠𝑖𝑡𝑒𝑠 𝐾 :𝐶𝑜𝑛𝑠𝑡𝑎𝑛𝑡 where it is understood that the “constant” K 1 , K 2 , K 3 may depend on the operating conditions, e.g. concentration of UH, current (𝐼), etc. Finally, it is assumed that the rate at which the concentration of active sites disappears (the sink of active sites - rate of loss of active sites and thus rate of 60 formation of inactive sites) is proportional to both the number of active sites and inactive sites (but with different rate constants, of course), i.e. |, ~, |= =−𝐾 N 𝐶 &% − 𝐾 " 𝐶 >% & (3.7) 𝐶 &%,%>kx :𝑆𝑖𝑛𝑘 𝑜𝑓 𝑎𝑐𝑡𝑖𝑣𝑒 𝑠𝑖𝑡𝑒𝑠 𝐾 N ,𝐾 " :𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡𝑠 In addition to the typical term proportional to the concentration of active sites (𝐾 N 𝐶 &% ), there is an term (𝐾 " 𝐶 >% & ) proportional to the number of inactive sites because without this term, the concentration of active sites never approaches zero (i.e., the cell never extinguishes) but instead approaches a constant value as shown in figure 3.19, which is not consistent with the experimental results. This additional term 𝐾 " 𝐶 >% & indicates a self-accelerating process which might exist if, for example, once an inactive site is formed, there is a higher rate of formation of inactive sites adjacent to the existing inactive site than on active sites. A possible explanation for this would be the formation of a polymer starting from a “seed” of a single molecule of propane that remains attached to the anode for an extended period of time without reacting. While the extinguishment can occur even with a = 1, the extinguishment is more abrupt (and thus closer to the experimental observation) if a > 1. Physically this might correspond to a multi-body reaction, e.g. with a = 4 then four molecules are needed to cause polymerization. It is emphasized that this model is only being used as a means to test possible mechanims of extinguishment and it not intended to be quantitative or predictive at this stage. In addition, the 𝐾 " term in 𝐾 " 𝐶 >% & is dependent on the concentration of ethylene in the fuel and can be expressed as: 𝐾 " = [[oo j@" (3.8) 𝑥:𝐶𝑜𝑛𝑐𝑒𝑛𝑡𝑟𝑎𝑡𝑖𝑜𝑛 𝑜𝑓 𝑒𝑡ℎ𝑦𝑙𝑒𝑛𝑒 𝑖𝑛 𝑝𝑝𝑚 61 With this simple model, we can express the total rate of change of formation of active sites as |, ~,~ |= = 𝐾 𝐶 >% −𝐾 N 𝐶 &% − 𝐾 " 𝐶 >% & (3.9) arbiturary constants were chosen to simulate the experimental data with assumption that all sites are active sites prior to testing, thus intial condition is 𝐶 &% =1 and 𝐶 >% =0 at time = 0. Figure 3.19. The direct propane PEMFCs extinguishment model based on mathematical derivation with constants shown in the figure (K 1 = 2.4 & K 2 = 2). (The 𝐾 " 𝐶 $% & term leads to self-accelerating of the rate of formation of inactive sites which is required to obtain a prediction of extinction; see text). 0 0.2 0.4 0.6 0.8 1 1.2 0 10 20 30 40 50 60 70 80 90 100 Concentration of Active Sites Time (Seconds) ExperimentalData Model K 3 = 9.5, a = 2.5 ModelK 3 = 4, a = 1 Model K 3 = 0 62 As shown in figure 3.19, the model without the self-accelerating term (𝐾 " 𝐶 >% & ) is inconsistent with the experimental data in that no extinction occurs and 𝐶 &% approaches a constant. However, for K 3 ≠ 0 the model behaves similarly to the experimental behavior in that abrupt extinguishment is predicted. More research and work is still needed to clearly identify intermediate species formed during oxidation of propane which might be responsible for more strongly adsorbed on platinum surface than that of CO. Figure 3.20. The direct propane PEMFC extinguishment models based on mathematical derivation with constants K 1 =2.4, K 2 = 2, a = 2.5, K 3 = 6600 / (x+737) with x = 0. 63 Figure 3.21. The direct propane PEMFC extinguishment models based on mathematical derivation with constants K 1 =2.4, K 2 = 2, a = 2.5, K 3 = 6600 / (x+737) with x = 1040. Figure 3.22. The direct propane PEMFC extinguishment models based on mathematical derivation with constants K 1 =2.4, K 2 = 2, a = 2.5, K 3 = 6600 / (x+737) with x = 1540. 64 Figure 3.23. The direct propane PEMFC extinguishment models based on mathematical derivation with constants K 1 =2.4, K 2 = 2, a = 2.5, K 3 = 6600 / (x+737) with x = 2000. The model is also in a good agreement with the experimental data for different concentration of ethylene gas in the fuel stream as shown in figures 3.20 ~ 3.23. The x variable in 𝐾 " constant was changed to the concentration of the ethylene gas (ppm) in the fuel stream and out of four different concentration tested, the model is in a good agreement with the data. The concentration of the UH affects the variable because it is believed that the 𝐾 " variable is responsible for polymerization of the species on the platinum surface, but the concentration of UH is helping to mitigate the extinguishment and increasing the longevity. 65 3.2. Testing Parameters Besides from the dynamics of direct hydrocarbon PEMFCs, testing parameters were explored to improve the performance of the fuel cell. The purpose of this study is to optimize the temperature, membrane thickness, catalyst loadings, and fuel cell apparatus flow height to maximize the performance at low temperatures. 3.2.1. Effect of Temperature Even though the electrochemical reaction rates of hydrocarbons on platinum catalyst at low temperatures are sluggish compared to those of hydrogen and methanol, it was observed that the direct propane PEMFCs did generate some power even at a temperature as low as 73˚C. It has shown that the performance of the fuel cell continually increased as the temperature increased as shown in figure 3.24. Once the temperature exceeded 114˚C, the fuel cell immediately extinguishes and the OCV exponentially decreased to zero. 66 Figure 3.24. Polarization curves of direct propane PEMFC at different operating temperatures. The flow rate was 1.2 L/min and 0.8 L/min for propane and oxygen, respectively. The scan rate was .133 mA/cm 2 s. The most optimal working temperature for Dupont TM Nafion® is around ~ 80˚C and this can be explained by findings reported by Lee. He claims that the ionic conductivities of Dupont TM Nafion® 112, 115, and 117 increased at elevated temperature from 30˚C to 80˚C, but slowly started to decrease in proton conductivity when the temperature was over 100˚C [31]. This is caused by dehydration of membranes due to high vaporization rate from Dupont TM Nafion® membranes at temperature above 100˚C at atmospheric pressure causing significant decrease in conductivity. It is also well understood that as the operating temperature increases, the theoretical OCV, elastic modulus, and proportional limit stress decrease at high temperature [27, 32]. It has been reported 0 2 4 6 8 10 12 14 16 18 20 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 10 20 30 40 50 60 70 Power Density (mW/cm 2 ) Voltage (V) Current Density (mA/cm 2 ) 73˚C 79˚C 86˚C 99˚C 107˚C 114˚C 67 by Xia et al. that the fatigue lifetime of the membrane strongly depends on the temperature because it can enhance effects of lowering break stress and raising break strain [32]. It was also reported by Xia et al. that the depth recovery ratio of Dupont TM Nafion® N-117, under the range of temperatures from 10˚C to 70˚C, decreases as temperature increases. Ohmic losses (resistance to flow of ions in the electrolyte) at high temperature is another concern. It is also studied by other scholars that in between temperature of 120˚C and 140˚C, the Dupont TM Nafion® membranes undergo glass phase transition, which may explain the sudden extinction of the fuel cell operating at over 114˚C [33-35]. Despite of many disadvantages of operating at higher temperature above 100˚C, the increase in the power density seen for the propane PEMFCs can be explained by the fact that high operating temperature (~400˚C) is ideal for propane oxidation due high chemical reaction and significantly depress the poisoning effect associated with strong adsorbed intermediates as seen in solid oxide fuel cells [15]. However, even operating at ~150˚C can greatly reduce the effect of CO poisoning and increase the chemical kinetics of electro-oxidation of propane [35]. This suggest that for the direct propane PEMFCs, the effects of propane reaction rates and ability to reduce CO poisoning might be more dominant than the effects of membrane dehydration, proton conductivity, ohmic losses, and durability at ~100˚C. The propane oxidation benefits from high operating temperature and can increase the power density, but due to mechanical properties of Dupont TM Nafion® membrane, it is important to optimize the ideal operating temperature by considering both chemical kinetics of propane and mechanical properties of Dupont TM Nafion® membrane. It is also of interest to test with PBI membranes which are known to have high glass phase transition temperature (~427˚C). 68 In fact, figure 3.24 can be redrawn as an Arrhenius plot (ln(power) vs. 1/T) as shown in figure 3.25. From the graph, the slope of the graph can be used to obtain effective activation energy (𝑠𝑙𝑜𝑝𝑒 = − t ~ < ; E a = effective activation energy & R = gas constant) of propane and it was calculated to be 5.2 kcal/mol. By comparison, Ahn et al. reported a similar value of 6.4 kcal/mol for the effective activation energy of propane oxidation on NH 3 -treated Pt catalyst in a heat- recirculating combustor [14]. Figure 3.25. Arrhenius plot based on the experimental result of temperature effect. 3.2.2 Effect of Membrane Thicknesses Membrane thickness can have a significant effect on the fuel cell’s performance for hydrogen PEMFCs. Since direct hydrocarbon PEMFCs are not well understood, three different commercially available membranes were studied: N-212 (50.8 micrometers), N-117 (183 y = -‐2617.3x + 9.6417 2 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 3 0.00255 0.0026 0.00265 0.0027 0.00275 0.0028 0.00285 0.0029 0.00295 ln(power) 1/T 69 micrometers), and N-1100 (254 micrometers). All three membranes were made out of the same material but have different thicknesses. Figure 3.26. Polarization curves of direct propane fuel cell with various membrane thicknesses at 80˚C. The flow rate was 1.2 L/min and 0.8 L/min for propane and oxygen, respectively. The scan rate was .133 mA/cm 2 s. The result shows that the maximum power density for N-1100, N-117, and N-212 was 12.62 mW/cm 2 , 12.54 mW/cm 2 , and 12.25 mW/cm 2 , respectively. As shown in figure 3.26, all three membranes display similar power density curves. For hydrogen PEMFCs, it has been generally understood that thinner membranes are used for high efficiency because of less resistance to flow protons from anode to cathode whereas thicker membranes reduce the crossover effect of hydrogen diffusion through the membrane. However, for methanol PEMFCs, thicker membrane might be beneficial because fuel crossover is a major challenge. For the propane PEMFCs all three 0 2 4 6 8 10 12 14 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0 10 20 30 40 50 Power Density (mW/cm 2 ) Voltage (V) Current Density (mA/cm 2 ) Nafion N-‐117 (7.2 mil, 183 micrometers) Nafion N-‐1110 (10 mil, 254 micrometers) Nafion N-‐212 (2 mil, 50.8 micrometers) 70 membranes yield similar results because transport of protons from anode to cathode is not an issue (low reaction rate) and the fuel crossover is not significant across the membrane. 3.2.3. Effect of Catalyst Loadings The electro-oxidation of propane was studied with regular platinum black as catalyst for two different loadings. The maximum power densities for high loading (anode: 15 mg/cm 2 & cathode: 8 mg/cm 2 ) and low loading (anode: 4 mg/cm 2 & cathode: 2 mg/cm 2 ) were 12.5 mW/cm 2 and 4.04 mW/cm 2 , respectively as shown in figure 3.27. The starting OCV is about ~ 0.85V for both loadings but it was observed that OCV slowly decreased over time to ~ 0.8V and ~ 0.65V for the high loading and low loading, respectively. The maximum current density for low loading fuel cell was at 13.4 mA/cm 2 compared to high loading fuel cell at 45.3 mA/cm 2 . It is important to note that there is a linear relationship between catalyst loading and power and current densities. By increasing the catalyst loading by 3.75 times (anode), the power density increased by 3.1 times and the current density increased by 3.38 times. The increase in power density and current density can be explained by the fact that with increase in catalyst loading, there are more active catalyst sites for propane oxidation and more tolerance to catalyst poisoning, which will be discussed later in this chapter. 71 Figure 3.27. Polarization curves of direct propane PEMFCs with low and high catalyst loadings at 80˚C. The flow rate was 1.2 L/min and 0.8 L/min for propane and oxygen, respectively. The scan rate was .133 mA/cm 2 s. Although the typical catalyst loading is about 0.1 mg/cm 2 [7], our study focuses on the maximum power energy output we can harvest from propane. Higher catalyst loading than 15 mg/cm 2 is also our interest of study, but there are limitations to direct painting process. As catalyst layers get too thick, more catalyst would come off from the electrode by the paint brush rather than being painted on and long painting process as loading increases is another concern. 3.2.4 Effect of Shear Stress It was previously found that the flow rates affect the propane fuel cell performance, in particular higher flow rates increased the time before the extinguishment occurred, but did not -‐1 1 3 5 7 9 11 13 15 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 10 20 30 40 50 Power Density (mW/cm 2 ) Voltage (V) Current Density (mA/cm 2 ) High Loading (Anode: 15 mg/cm^2 Cathode: 8 mg/cm^2) Low Loading (Anode: 4 mg/cm^2 Cathode: 2 mg/cm^2) 72 have a significant effect on the power before the extinguishment. Increasing flow rate is equivalent to increasing the shear stress and it is predicted that the shear stress might help to mitigate CO adsorption on the platinum catalyst. The equation of bulk velocity and shear stress is given by, Bulk Velocity 𝑈= (3.10) 𝑈:𝐵𝑢𝑙𝑘 𝑉𝑒𝑙𝑜𝑐𝑖𝑡𝑦 𝑉: 𝑉𝑜𝑙𝑢𝑚𝑒 𝐹𝑙𝑜𝑤 𝑅𝑎𝑡𝑒 𝐴 u :𝐶𝑟𝑜𝑠𝑠 𝑆𝑒𝑐𝑡𝑖𝑜𝑛𝑎𝑙 𝐴𝑟𝑒𝑎 Shear Stress 𝜏 = ~ n (3.11) 𝐹 y :𝐹𝑜𝑟𝑐𝑒 𝑃𝑎𝑟𝑎𝑙𝑙𝑒𝑙 𝑡𝑜 𝑆𝑢𝑟𝑓𝑎𝑐𝑒 𝐴 % :𝐴𝑟𝑒𝑎 𝑜𝑓 𝑡ℎ𝑒 𝑆𝑢𝑟𝑓𝑎𝑐𝑒 ℎ:𝐷𝑒𝑝𝑡ℎ 𝑜𝑓 𝑡ℎ𝑒 𝐹𝑙𝑜𝑤 𝐶ℎ𝑎𝑛𝑛𝑒𝑙 It may be possible that by increasing shear stress by a significant amount, adsorbed species such as CO might be stripped from the platinum surface. Although there are no literatures to support this theory, some preliminary work by Jiyuan et al. at CPL has shown that by inserting non-reactive plastic into flow channels (thus decreasing the height and increase in shear stress), the time for extinct of the direct propane PEMFC nearly doubled as shown in figure 3.28. 73 Figure 3.28. The power density curve of the same fuel cell: one with decreased flow channel volume and one without. Both tests were run on a hydrazine treated propane PEMFCs run at a constant current of 36 mA/cm 2 . However, this study was not rigorously and systematically studied and needs to be quantified. The report by Jiyuan, failed to report what kind of plastic sleeves were used and how they were secured to the fuel cell flow field to prevent it from moving during the experiment. Additionally, no information was given on the thickness of the plastic sleeves, therefore the height of the decreased flow channel is unknown. For future study, new fuel cells will be machined at the machine shop using the same type of graphite plates (impervious bipolar graphite plates FC- GR347B) and configuration, but only varying the height of the flow field (200% & 50% of the current fuel cell flow field height). This method will be more systematic and eliminates any variables that might have been introduced in Jiyuan’s experiment. 0 2 4 6 8 10 12 14 16 18 20 0 50 100 150 200 250 300 350 400 450 500 Power Density (mW/cm 2 ) Time (seconds) Normal Flow Channel Height Decreased Flow Channel Height 74 At the USC machine shop, two other fuel cells were fabricated with different heights for the flow channel height with all other specifications of the graphite plate and testing parameters staying constant. The normal flow channel height of the fuel cell was ~ 12 mm and the height was halved and doubled to increase the shear stress by a factor of 2 and decrease the shear stress by a factor of 2, respectively. As shown in figure 3.29, contrary to previous result, there was no effect of the flow channel height (shear stress) on the performance of the fuel cell. The only different testing parameter between figure 3.28 and figure 3.29 is that in figure 3.28, the platinum was pretreated with hydrazine. It was unfortunate, but the same result with hydrazine treatment of platinum did not work (will be discussed in chapter 4), so the exact same testing condition was not possible. Figure 3.29. Galvanostatic mode of direct propane PEMFCs at 24 mA/cm 2 at three different heights without hydrazine treatment. 0 5 10 15 20 25 0 20 40 60 80 100 120 Power Density (mW/cm 2 ) Time (Seconds) Normal Height Half Height Double Height 75 The flow rate of the fuel was also halved and doubled to halve and double the bulk velocity, thus decreasing the shear stress by a factor of 2 and increasing the shear stress by a factor of 2, respectively. Even with the change in the bulk velocity, the performance of the fuel cell was not affected. Figure 3.30. Galvanostatic mode of direct propane PEMFCs at 36 mA/cm 2 at three different fuel flow rate at the normal height without hydrazine treatment. The working temperature was at 80˚C. Previous results showed a significant effect of shear stress on the performance of PEMFCs but the same result was not repeatable. In addition, previous result reported that higher flow rate also improved the performance of the fuel cell, but that was also not the case as shown in figure 3.30. The only explanation possible for increase in the performance with increase in flow rate is because of increase in temperature. The previous experimental setup did not have heating pads 0 5 10 15 20 25 30 35 0 10 20 30 40 50 60 Power Density (mW/cm 2 ) Time (Seconds) 1200 mL/min 600 mL/min 2400 mL/min 76 attached to the fuel cell, therefore the working temperature was dependent on the flow rate of humidified gas (1.2 L/min for anode & 0.8 L/min for cathode is the flow rate needed to reach 80˚C). By increasing the flow rate, it is suspected that the working temperature of the fuel cell increased as well. 3.2.5 Effect of the Fuel Type For portable devices, it would be beneficial to use butane as fuel instead of propane because butane is in liquid state at ≈ 2 atmosphere at room temperature as opposed to propane at ≈ 12 atmosphere. Therefore, the performance of isobutene and n-butane as fuel for PEMFCs at low temperature (80˚C) was also explored. It can be seen from figure 3.31 that n-butane exhibit similar performance to that of propane, but isobutene shows slightly lower performance and all hydrocarbons show the sudden extinction behavior at high currents as shown in figure 3.32. 77 Figure 3.31. Polarization curves of direct propane n-butane, and isobutane PEMFCS scanning at 0.133 mA/cm 2 s and operating at 80˚C. Figure 3.32. Galvanostatic mode of direct propane, n-butane, and isobutene PEMFCs at 36 mA/cm 2 and operating at 80˚C. The slight lower performance of n-butane might be because there is one more carbon atom in the chain and there are more intermediate steps to completely oxidize to CO 2 , thus requires more energy and higher temperature to exhibit the same power density as that of propane. The discrepancy between n-butane and isobutene is because they have different molecular structure. The n-butane has a long straight chain of carbon atoms, whereas isobutene has a carbon attached to 3 other carbon atoms (cyclo-shape). Since long straight chain of carbon atoms has more surface area to react at the catalyst surface as opposed to a ball shape molecule, n-butane exhibit higher power density than that of isobutene. Another explanation is that n-butane has 6 primary C-H 0 5 10 15 20 25 30 35 1 10 100 1000 Power Density (mW/cm 2 ) Time (Seconds) Propane N-‐butane Isobutane 78 bonds and 4 secondary C-H bonds and isobutene has 9 primary C-H bonds and 1 secondary C-H bonds. Since it is easier to break secondary C-H bonds, it will easier to oxidize n-butane compared to that of isobutane, thus showing better performance. The feasibility of using unsaturated hydrocarbon as fuel for low temperature PEMFCs also has been explored. Ethylene was used as fuel because it was the best additive tested so far. As shown in figure 3.33, the fuel cell was able to carry almost three times more current density compared to that of propane (34 mA/cm 2 vs. 90 mA/cm 2 ) and showed similar maximum power density (≈ 8.5 mW/cm 2 ). Figure 3.33. Polarization curve of direct ethylene PEMFC scanning at 0.133 mA/cm 2 s. The operating temperature was held constant at 80˚C and the same experimental setup and testing parameters (flow rate, catalyst loadings…etc.) were employed. Using ethylene as fuel, the dynamics of the fuel was also tested by holding the current at various currents for 1000 seconds. As shown in figure 3.34, it was interesting to note that there was no sudden extinguishment behavior observed at high currents and rather an exponential decay 79 of the performance, which is the expected behavior for PEMFCs. Unlike direct hydrocarbon PEMFCs, ethylene PEMFCs is not time dependent. Figure 3.34. Galvanostatic mode at various currents for direct ethylene PEMFC at 80˚C. The operating temperature was held constant at 80˚C and the same experimental setup and testing parameters (flow rate, catalyst loadings…etc.) were employed. However, using ethylene as fuel for portable devices is not favorable compared to hydrocarbons because critical pressure for ethylene is ~50 atm which is 25 times higher than that of n-butane. Instead, ethylene glycol (C 2 H 6 O 2 ), viscous liquid, might be an alternative fuel to power portable devices. 80 3.2.6 Aluminum Bipolar Plates In the beginning of this project, a new fuel cell apparatus was needed for testing. Due to high cost of commercially available fuel cell apparatus ($3,000~$5,000), making a cheaper fuel cell apparatus using aluminum was attempted at the USC machine shop. Metallic alloys are particularly attractive due to low manufacturing cost, high-volume manufacturing availability, high thermal and electrical conductivities, and impervious to fluid. Since graphite plates are porous, commercially available fuel cell test cells go through a coating process to make them impervious. Aluminum plates were machined at the USC machine shop and the total cost was about $200-$300. Figure 3.35. The bipolar plate for PEMFC application with active area 5 cm 2 and pin post type flow field was created in SolidWorks. 81 Figure 3.36. The complete aluminum fuel cell apparatus after being machined in the USC machine shop. The dimensions of the fuel cell apparatus (borrowed from Prakash Group) were carefully measured with a caliper and it was created in SolidWorks as shown in figure 3.35. The active area was 5 cm 2 with pin post type flow field. The fuel cell apparatus was then machined at the machine shop by Mike using aluminum instead of graphite as shown in figure 3.36. 82 Figure 3.37. The aluminum bipolar plates of the fuel cell after a couple hours of testing. Figure 3.38. A closer view of the aluminum plate after a couple hours of testing. 83 The fuel cell was first tested with hydrogen to assure that the performance was similar to that of graphite plates. The results were compared with commercially bought fuel cell apparatus (Prakash Group). In the beginning, both machined aluminum fuel cell and commercially bought graphite fuel cell showed similar results. However as shown in figure 3.37 and figure 3.38, only after a couple hours of testing, we can see aluminum bipolar plates corroding. This is expected because of the hygroscopic nature of the sulfonic acid group in the Nafion® membranes. Since the operating conditions inside the fuel cell is acidic, the aluminum will corrode and degrade in performance over time. Aluminum is also prone to oxidation when in contact with water and will lead to increase in electronic resistivity, contact resistances, and uneven current distribution. Because of these challenges, coating of metallic surface is necessary. However, additional cost and pinhole defects leading to local corrosions are another concerns. Consequently, graphite is the most widely used material in the fuel cell industry. Graphite has high resistance to corrosion in a wide oxidizing potential range, chemically stable, light and conducts adequate electricity and heat. The fuel cell apparatus was re-machined at the USC machine shop using high density isomolded, impervious, and resin impregnated graphite (FC-GR347B). 84 3.3 Characterization Results Characterization instruments were utilized to study the platinum catalyst before and after the fuel cell experiment in hopes to discover any explanation for the sudden extinction behavior of direct propane PEMFCs. XRD was used to study any alteration in crystal structures and atom spacing, XPS was used to study chemical and electronic state of the catalyst, SEM to take image of surface of the catalyst. 3.3.1 X-Ray Photoelectron (XPS), X-Ray Diffraction (XRD), and Scanning Electron Microscope (SEM) First, XPS of new platinum catalyst and platinum catalyst after fuel cell testing is compared. From the XPS data shown in figure 3.39 & figure 3.40, there is a small concentration of platinum oxide in both catalyst and the concentration of platinum oxide is higher for that of platinum catalyst after electrocatalysis. During the fuel cell testing, it is possible that oxygen from the cathode crossed over to form platinum oxide. Increase in the platinum oxide will decrease the performance of the fuel cell because there is less platinum metal, therefore less electrochemical active surface area. The XPS machine used in this study is Kratos Axis Ultra from Center of Excellence in Nano Imaging (CNI). The pressure at the sample analysis chamber was x -10 Torr and X-ray gun used in this study is Mono Aluminum X-ray with power settings 6 mA and 10 kW. 85 Figure 3.39. XPS (Kratos Axis Ultra) result of new platinum catalyst with Mono Aluminum X-ray working at x -10 Torr, power setting: 6 mA and 10 kW. 86 Figure 3.40. XPS (Kratos Axis Ultra) result of platinum catalyst after electrocatalysis with Mono Aluminum X-ray working at x -10 Torr, power setting: 6 mA and 10 kW. The platinum metal and platinum oxide curves were fitted based on table 3.3. The first curve shown in the XPS data is the 4f5/2 orbital and the second curve is the 4f7/2 orbital. The compound value for binding energy can be computed by adding 3.33 eV to the table because Pt 4f7/2 – 4f5/2 splitting is 3.33. eV. 87 Table 3.3. Platinum binding energy for 4f7/2 values [47]. For future study, it will be also interesting to do XPS testing on the hydrazine treated platinum black. As it will be discussed in the next chapter, hydrazine treated platinum has shown a significant improvement on the performance of direct propane PEMFCs. It is known that hydrazine is a strong reducing agent, and it is speculated that platinum oxide might be getting reduced to platinum metal. Thus, there will be more active sites for oxidation reaction for both hydrocarbon and as well as hydrogen. XRD testing of the catalyst has been studied to see if there is any change in the crystalline structure of the catalyst before and after testing. As shown in figure 3.41, both new and used platinum catalyst shown exactly the same spectrum. It can be assumed the crystalline structure of the catalyst did not change over course of testing. As shown in the figure, the first reflection at 39.7˚ attribute to platinum [111] plane, 46.4˚ to [200] plane, and 67.6˚ to [220] plane. The XRD used in this experiment is Rigaku Ultima IV using 𝜅−𝛼 X-ray working at 40 kV, 44 mA, and 1.76 kW. 88 Figure 3.41. XRD (Rigaku Ultima IV) result of new platinum catalyst and platinum after electrocatalysis using 𝜅−𝛼 X-ray working at 40 kV, 44 mA, and 1.76 kW. Lastly, images of the catalyst were taken by employing SEM. The surface image of the catalyst was taken at three different magnifications: low, medium, and high. As seen at all three different magnifications, after the platinum catalyst has been tested in the fuel cell, it seems like there is some agglomeration of the catalyst as shown in figure 3.42. The SEM used in this experiment is JEOL 7001 low vacuum with voltage set at 15 kV, probe current at 12 mA and working distance at 10 mm. 0 2000 4000 6000 8000 10000 12000 0 10 20 30 40 50 60 70 80 90 100 Intensity (Counts per Second) 2 Theta New Platinum Platinum after Electrocatalysis Platinum [111] Platinum[200] Platinum[220] 89 Figure 3.42. SEM (JSM-6610LV low vacuum) images of new platinum catalyst and platinum catalyst after electrocatalysis. Upper Left: New platinum at low magnification. Upper Right: Platinum after electrocatalysis at low magnification. Middle Left: New platinum at medium magnification. Middle Right: Platinum after electrocatalysis at medium magnification. Bottom 90 Left: New platinum at high magnification. Bottom Right: Platinum after electrocatalysis at high magnification. Due to limitations of SEM (only characteristic of the catalyst at surface level is shown), it is difficult to derive any meaningful information between the new catalyst and catalyst after electrocatalysis. Thus, it is of interest to characterize of the size and shape of nanoparticles of the catalysts under Transmission Electron Microscope (TEM). While SEM focuses only on the sample’s surface morphology of the sample, by employing TEM, it can provide the details about internal composition, crystallization, stress, and even magnetic domains. 91 Chapter 4 – Future Work 4.1. In Situ Spectroscopy Method More research is still needed to elucidate the cause of sudden extinguishment of the fuel cell. The ideal technique is to employ in situ methods of spectroscopy to study adsorption of molecules and intermediate species during oxidation of hydrocarbon molecules on the platinum surface under operando conditions, which can be a challenging task. So far, there are no literatures reported on directly using spectroscopy methods on operating fuel cells, however, there are literatures reported on using spectroscopy methods on half-cells. It is a great of interest to use in situ spectroscopy methods in fuel cells and batteries because a comprehensive understanding of electrode reactions is necessary to optimize electrode materials and study the nature of adsorbates and solution species in electrochemical reactions [48]. In situ spectroscopy also has been utilized for methanol, formaldehyde, and formic acid half cells for understanding the mechanism and kinetics of the reaction at the molecular scale. It has been identified that CO and formate are adsorbed on the electrode surface and formate is oxidized to carbon dioxide or desorbed to formic acid [49]. As reported by previous investigators, the same technique can be possibly used for direct hydrocarbon fuel cell to optimize the performance and study the kinetics and mechanisms of hydrocarbon oxidation at low temperature via platinum catalyst. The most popular technique is the in situ time-resolved Fourier-transform infrared spectroscopy (FTIR). In-situ time resolved FTIR can provide information about both structural and transient dynamic molecular structures. The key feature of time-resolved FTIR is that the spectral data is collected as a function of time instead of frequency. The conventional method of 92 spectroscopy is to use monochromatic light beams at a time and measure how much of the light is absorbed, however, for time-resolved FTIR, this technique shines a beam containing many frequencies of lights at once and measure how much light is absorbed by the sample over time. After, computer processing is required to turn the raw data (interferogram) into the desired result by inferring light absorption for each wavelength to light absorption for each mirror position, process known to be a Fourier transform. Figure 4.1. Schematic drawing of in-situ time resolved FTIR [50]. A successful experimental setup of in-situ time-resolved FTIR in the lab on half-cell can provide an insightful information about hydrocarbon oxidation in a similar fuel cell operating setting. Since no intermediate species are detected through GC and IR, it might be possible that 93 the intermediate species are strongly adsorbed on the platinum surface or there might be a very small ppm amount of intermediate species causing the extinguishment and somehow completely oxidizes as the current is shut off. This technique can potentially identify the intermediate species in real time during the oxidation of hydrocarbons and hopefully understand the extinguishment behavior of direct hydrocarbon PEMFCs. 4.2. Platinum Treatment - Hydrazine Platinum black has been the most favored catalyst for PEMFCs for decades, but limited availability of platinum and high cost of platinum (~ $200 per gram) have been potential threats to commercialization of fuel cells. Many research efforts have been made to shift away from platinum and search for alternative catalyst with heat treatments to increase the surface area or lower the activation energy [51, 52]. Works by Jaouen et al. suggest that heat treated Fe/N/C and Co/N/C catalysts with ammonia have been successful in increasing the site density for oxygen reduction reaction [53]. The steps of heat treatment of the catalyst follow; i) incorporation of nitrogen atoms in the carbon ii) micropore formation through reaction between ammonia and carbon iii) competition of active sites in the micropores by reaction of iron with ammonia. These works strongly suggest that formation of micropores increases the site density of such catalysts. Earlier work by McCabe et al. studied catalytic etching of platinum with heated NH 3 -air mixtures and suggest that etching produces either flat facets, curved surfaces, pitting, or a combination of several modes [54]. In accordance with Jaouen et al., micropores and stepped planes on on platinum wires were also observed after ammonia heat treatment which might suggest that heat treatment of platinum black might be promising for the direct propane PEMFCs. Similar 94 results have been observed in our lab, when Hong et al. performed SEM images on Pt foil before the ammonia treatment and after as shown in figure 4.2. Figure 4.2. Scanning electron microscopy (SEM) pictures of platinum foil before (left) and after heating in NH 3 (right) – air mixture for an hour. The bare Pt foil was burned with propane–air mixtures with ≈5% of the propane replaced by ammonia for one hour. (field of view 6µm x 8µm) [29]. While the exact mechanism behind formation of micropores and stepped planes during ammonia heat treatment are complicated and not fully understood, one tentative explanation might be that during the heat treatment process, the surface may be unstable leading to lower surface tension and metal migration forming pits and micropores in the structure. Additionally, due to limitation of SEM (JSM-6610LV low vacuum SEM), low resolution at micro scale, Transmission Electron Microscope (TEM) will be also considered to have a better understanding and image of micro scale of catalyst powders. A preliminary work by Hong et al. in 2011 have claimed success with ammonia and hydrazine treated platinum catalyst for the direct propane PEMFCs. First she showed that the catalyst surface area increased during the platinum treatment with ammonia and hydrazine using XRD results as shown in figure 4.3. 95 Figure 4.3. XRD patterns for the untreated Pt black and treated Pt catalyst [29]. With this data and using the Debye-Scherrer equation 𝑑 = o. ¡ ¢ £ (/¤) ¥¦§¨ ©~ª (4.1) where 𝜆 ¬ ¢ is the wavelength of the X-rays (1.54056 Å of Cu), θ is the angle at the peak maximum, and 𝛽 (N¨) is the width at half height of the diffraction peak. The surface area (SA), in m 2 /g, can also be calculated with the equation 𝐴= [ooo ¯ ~ | (4.2) where d is the mean particle size in nm and 𝜌 u&= is the catalyst density, as 21.45 g/cm 3 for Pt as shown in table 4.1. Electrocatalyst Average particle size, d (nm) SA (m 2 /g) PtBl 10.56 26.49 Pt(NH 3 ) 9.39 29.79 Pt (N 2 H 4 ) 7.68 36.42 Table 4.1 Average particle size and SA of untreated Pt black and treated Pt catalysts [29]. 96 It is clear from her results that the surface area of catalyst increased, which is a promising result for all types of PEMFCs. However, it is also important to consider platinum oxide compounds in the catalyst. This increase in the surface area might be due the fact that the oxygen atom might be getting reduced with hydrazine, creating more platinum active sites and reducing the average particle size. She further supports her results with cyclic voltammograms (CVs) results to show that the electrochemical active surface area (ECSA) increased with hydrazine treated platinum black as shown in figure 4.4 and limiting current density significantly increase as shown in figure 4.5 Figure 4.4. Cyclic voltammograms of PtBl and Pt(N 2 H 4 ) catalysts. Measurements were performed in 0.5 M H 2 SO 4 solution saturated by Ar. Scan rate: 20 mV/s [29]. -0.05 -0.04 -0.03 -0.02 -0.01 0.00 0.01 0.02 0.03 0.04 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 j (mA/cm 2 ) E (V) vs. RHE PtBl Pt(N2H4) 97 Figure 4.5. Cyclic voltammograms of propane oxidation on PtBl and Pt(N 2 H 4 ) catalysts. Measurements were performed in 0.5 M H 2 SO 4 solution under propane flow at 90˚C. Scan rate: 20 mV/s [29]. At last, in the cell performance, it was shown that the platinum black with ammonia showed a small improvement, but was able to almost the double power density by treating platinum with hydrazine as shown in figure 4.6. -0.08 -0.06 -0.04 -0.02 0.00 0.02 0.04 0.06 0.08 0.10 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 j (mA/cm 2 ) E (V) vs. RHE PtBl Pt(N2H4) 98 Figure 4.6. Polarization curves of direct propane PEMFC with various anode catalysts at 80°C. Propane pressure and flow rate: 1 atm and 1.2 L/min, oxygen pressure and flow rate: 1 atm and 0.5 L/min. Both streams were humidified at 90˚C [29]. Numerous attempts have been made to create hydrazine treated platinum black to duplicate her result, but the result was not repeatable. The procedure for hydrazine treatment of platinum black follows: 1. 0.6g platinum black was added into the 100mL 25% hydrazine aqueous solution (anhydrous 98%, Sigma-Aldrich) with continuous stirring 2. The solution was heated to 40˚C and kept at this temperature for 4 hours. 3. The reaction mixture was heated to 100˚C for one hour 4. After the solution was cool down to room temperature, the treated platinum was washed with de-ionized Milipore water and naturally dried before being used as fuel cell catalysts. 0 4 8 12 16 20 24 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 10 20 30 40 50 60 70 80 90 Power Density (mW/cm 2 ) Voltage (V) Current Density (mA/cm 2 ) PtBl Pt(NH3) Pt(N2H4) 99 Even after several attempts of exactly following the procedure written in the report, the performance of the direct propane PEMFCs yields similar result to that of non-treated platinum black or sometimes the result was worse. One notable observation of the platinum catalyst treatment with hydrazine was serious agglomeration of catalyst as shown in figure 4.4. Figure 4.7. The platinum catalyst with hydrazine treatment process recommended by Hong et al. As shown, serious agglomeration of the catalyst. As shown in figure 4.7, serious agglomeration of the catalyst has been observed after the treatment process. Because of serious agglomeration, the catalyst had to be broken down with a mortar and pestle before the ink preparation. The treatment process of platinum black hydrazine is still work in progress and other treatment methods of hydrazine treatment will be explored. 100 4.3. Catalyst Synthesis To successfully commercialize direct hydrocarbon PEMFCs, it is important to reduce the platinum loading. One simple way to significantly lower the loading without losing much of the performance is to employ platinum on carbon catalyst. Since the platinum particles are supported on carbon, it allows to maximize the electrochemical surface area of the catalyst. Further investigation of catalyst synthesis for direct hydrocarbon PEMFCs is necessary to optimize such as catalyst activity, selectivity, stability, and cost. 101 Chapter 5 – Conclusion This work successfully demonstrated working direct hydrocarbon PEMFCs despite low reactivity of hydrocarbons at low temperatures (< 100˚C) compared to those of methanol and hydrogen. The performance of the fuel cell is found to be time dependent; faster scan rates produce higher maximum power densities and higher currents yield faster extinguishment time. The extinguishment of the fuel cell can be avoided by operating the fuel cell in the load-interrupt mode where the load is applied on and off. The working hypothesis is that turning off the load somehow desorbs any poisoning species on platinum surfaces. By employing this operating mode, the average power density can be 1.5 times higher than that of galvanostatic mode. The extinguishment of the fuel cell can also be prolonged by bleeding small amount of UH (≈ 0.2%) into the anode fuel stream. It is suspected that UH is retarding the polymerization process of some species on the platinum surface. Out of four UHs tested, ethylene seems to yield the best result and carbon monoxide, a well known specie for catalyst poisoning, did not contribute to the extinguishment of the fuel cell. The GC and IR analysis of the anode fuel exhaust suggest that only carbon dioxide and unreacted fuel was detected as major species and methane and ethane as minor species. No other intermediate species were detected both in gas and liquid phase. It was found that at lower currents, there is enough time for inactive sites to convert back to active sites, but as current increases, there is not enough time for inactive sites to convert back to active sites leading to low carbon conversion ratio. It was also interesting to note that almost 20 electrons reacted per propane molecule at every current, which is a promising result. These experimental data were taken into a simple mathematical model in order to understand the behavior of direct hydrocarbon PEMFCs at a low temperature. It has been concluded from the model that the sudden extinguishment of the 102 fuel cell can be best described by polymerization of some species on the platinum surface. However, the polymerization process can be mitigated by operating at low currents, employing load-interrupt mode, or bleeding a small amount of UH into the fuel stream. More work and research is still needed to elucidate the mechanism of hydrocarbon oxidation at low temperatures in PEMFC working environment. One way to identify all intermediate species during hydrocarbon oxidation in real time is to employ in-situ time resolved FTIR on half cell experimental setup. A successful experimental setup can lead to better understanding of oxidation mechanism and hopefully avoid the extinguishment behavior and increase the performance. 103 References 1. Gaur S, Reed T, Thermal data for natural and synthetic fuels. Marcel Dekker, New York, 1998 2. S. R. Narayan and Thomas I. Valdez. High-Energy Portable Fuel Cell Power Sources. The Electrochemical Society Interface Winter 2008. 3. Hongting Pu, Polymers for PEM fuel Cells. Wiley, New Jersey, 2014. 4. Vladimir S. Bagotsky, Fuel Cells Problems and Solutions. Wiley, New Jersey, 2012. 5. M. M. Mench, H. M. Chance, and C. Y. Wang, Direct dimethyl ether polymer electrolyte fuel cells for portable applications. Journal of The Electrochemical Society, 151 (1, 2004), A144-A150 6. M. M. Mench, H. M. Chance, C. Y. Wang, Direct Dimethyl Ether Polymer Electrolyte Fuel Cells for Portable Applications. Journal of the Electrochemical Society, 151 (1) A144- A150 (2004). 7. G. A. Olah, A. Goeppert, G.K.S. Prakash, Beyond Oil and Gas: The Methanol Economy. Wiley, New York, 2009 8. S. Surampudi, S.R. Narayanan, E. Vamos, H.A. Frank, G. Halpert, G.A. Olah, G.K. Surya Prakash, US Patent No. 5,599,638 (1997) 9. S. Surampudi, S.R. Narayanan, E. Vamos, H.A. Frank, G. Halpert, G.K. Surya Prakash, US Patent No. 6,248,460 B1 (2001). 10. Trygve Riis, Elisabet F. Hagen, Preben J. S. Vie, Oystein Ulleberg, Hydrogen Production R&D: Prioriteis and Gaps. OECD/IEA, France 2006. 104 11. H. L. Chen, H. M. Lee, S. H. Chen, Y. Chao, M. B. Chang, Review of plasma catalysis on hydrocarbon reforming for hydrogen production—Interaction, integration, and prospects. Applied Catalysis B: Environmental Volume 85, Issues 1–2, 17 December 2008, Pages 1– 9 12. S. Ahmed, Hydrogen from hydrocarbon fuels for fuel cells. International Journal of Hydrogen Energy Volume 26, Issue 4, April 2001, Pages 291–301 13. O. Savadogo, F.J.R. Varela, Low-temperature direct propane polymer electrolyte membranes fuel cell (DPFC). J. New Mater. Electrochem. Sys., 4 (2001), pp.93-97 14. J. Ahn, C. Eastwood, L. Sitzki and P.D. Ronney, Gas-phase and catalytic combustion in heat-recirculating burners. Proceedings of the Combustion Institute 30, Issue 2 (2005), pp. 2463-2472 15. Z. Shao, S. Haile, J. Ahn, P. Ronney, Z. Zhan, & S. Barnett, A Thermally Self-Sustained Micro Solid-Oxide Fuel Cell with High Power Density. Nature (2005) Vol. 435, pp. 795- 798 16. A.C. Fernandez-Pello, Proc. Combust. Inst. 29 (2002) 883-899 17. L.W. Niedrach, The performance of hydrocarbons in ion exchange membrane fuel cells. Journal of The Electrochemical Society, 109, Issue (11) (1962), pp. 1092-1096 18. W.T. Grubb, C. J. Michalske, A High Performance Propane Fuel Cell Operating in the Temperature Range of 150–200˚C. Journal of The Electrochemical Society, 111, Issue (9) (1964), pp. 1015-1019 19. E.J. Cairns, Hydrocarbon Fuel Cells with Fluoride Electrolytes. Journal of The Electrochemical Society, 113, Issue (11) (1966), pp. 1200-1204 105 20. O. Savadogo, Emerging Membranes for Electrochemical Systems: (I) solid polymer electrolyte membranes for fuel cell systems. Journal of New Materials for Electrochemical Systems 1. 47-66 (1998). 21. O. Savadogo, B.Xing, Hydrogen/Oxygen Polymer Electrolyte Membrane Fuel Cell (PEMFC) based on acid-doped polybenzimidazole(PBI). Journal of New Material for Electrochemical Systems, 3, 345-349 (2000) 22. O. Savadogo, F.J.R. Varela, Low-temperature direct propane polymer electrolyte membranes fuel cell (DPFC). J. New Mater. Electrochem. Sys., 4 (2001), pp.93-97 23. O. Savadogo, F.J.R. Varela, Real time mass spectrometric analysis of the anode exhaust gases of a direct propane fuel cell. Journal of The Electrochemical Society, 152, Issue (9) (2005), pp. 1755-1762 24. F.J. Rodriguez Varela, O. Savadogo, The Effect of Anode Catalysts on the Behavior of Low Temperature Direct Propane Polymer Electrolyte Fuel Cells (DPFC). Journal of New Materials for Electrochemical Systems 9, 127-137 (2006) 25. W.S. Li, D.S. Lu, J.L. Luo and K. T. Chuang, Chemicals and energy co-generation from direct hydrocarbons/oxygen proton exchange membrane fuel cell. J. Power Sources, 145 (2005), pp.376-382 26. Frano Barbir, PEM Fuel Cells Theory and Pratice. Elsevier INC, Waltham 2013. 27. Fuel Cell, http://fuelcell.com/fuel-cell-hardware/ 28. Praxair Propane Catalog, https://www.praxair.com/- /media/corporate/praxairus/documents/specification-sheets-and-brochures/gases/propane- c3h8-spec-sheet-ss-p4646.pdf?la=en&rev=fca0eb32a0f34624a5e3d888bb1c0aba 29. Dan Hong, Studies on Direct Hydrocarbon Fuel Cells at Low Temperature, 2011. 106 30. T. Ioroi, N. Fujiwara, Z. Siroma, K. Yasuda, Y. Miyazaki, Platinum and Molybdenum Oxide Deposited Carbon Electrocatalyst for Oxidation of Hydrogen Containing Carbon Monoxide. Electrochemical Communications 4 (2002) 442-446 31. C. H. Lee, H. B. Park, Y. M. Lee and R. D. Lee. Ind. Eng. Chem. Res., 2005, 44 (20), pp 7617–7626 32. R. Xia, H. Zhou, R. Wu, W. Wu, Nanoindentation Investigation of Temperature Effects on the Mechanical Properties of Nafion 117. Polymers 2016, 8, 344 33. J. Zhang, Z. Xie, J. Zhang, Y. Tang, C. Song, T. Navessin, Z. Shi, D. Song, H. Wang, D. P. Wilkinson, Z. Liu, S. Holdcroft, High Temperature PEM Fuel Cells. Journal of Power Sources 160 (2006) 872-891 34. A. Cbandan, M. Hattenberger, A. El-kharouf, S. Du, A. Dhir, V. Self, B. G. Pollet, A. Ingram, W. Bujalski, High Temperature (HT) polymer electrolyte Membrane Fuel Cells (PEMFC) – A Review. Journal of Power Sources 231 (2013) 264-278 35. S. S. Araya, F. Zhou, V. Liso, S. L. Sahlin, J. R. Vang, S. Thomas, X. Gao, C. Jeppesen, S. K. Kaer, A Comprehensive Review of PBI-based High Temperature PEM Fuel Cells. International Journal of Hydrogen Energy 41 (2016) 21310-21344 36. J. H. Yoo, H. Choi, C. Chung, S. M. Cho, Fuel Cells Using Dimethyl Ether. Journal of Power Sources Volume 163, Issue 1, 2006 (103-106) 37. J. O. Jensen, A. Vassliev, M. I. Olsen, Q.F. Li, C. Pan, L.N. Cleemann, T. Steenberg, H.A. Hjuler, N. J. Bjerrum, Direct Dimethyl Ether Fueling of a High Temperature Polymer Fuel Cell. Journal of Power Sources Volume 211, (2012) 173-176 38. S. Vafaeyan, A. St-Amant, M. Ternan, Propane Fuel Cells: Selectivity for Partial or Complete Reaction. Journal of Fuels Volume 2014 Article ID 485045 107 39. M.G. Sustersic, R. Cordova, W. E. Triaca, A. J. Arvia. The electrosorption of Methane and Its Potentiodynamic Electrooxidation on Platinzed Platinum Division Electroquimica, 1990 40. S. Y. Hsieh, K. M. Chen, Anodic Oxidation of Methane. Electrochemical Science and Technology Vol. 124 No. 8 (1977) 1171-1174 41. V.S. Bagotzky, Y. B. Vassiliev, O. A. Khazova, Generalized Scheme of Chemisorption, Electrooxidation and Electroreduction of Simple Organic Compounds on Platinum Group Metals. Journal of Power Source 81 (1977) 229-238 42. J. O’M. Bockris, E. Gileadi, G. E. Stoner, The Anodic Oxidation of Saturated Hydrocarbons. The Journal of Physical Chemistry Volume 73 (1969) 427-434 43. N. Zamel, X. Li, Transient Analysis of Carbon Monoxide Poisoning and Oxygen Bleeding in a PEM Fuel Cell and Anode Catalyst Layer. International Journal of Hydrogen Energy 33 (2008) 1335-1344 44. Gottesfeld US Patent No. 4,910,099 (1990). 45. Engel T, Ertle G. Oxidation of Carbon Monoxide. In: King D, Woodruff D, editors. The Chemical Physics of Solid Substances and Heterogeneous Catalysis, vol. 4: fundamental studies of heterogeneous catalysis. Amsterdam: Elsevier; 1982. P73-93 46. Norton P. In: King D, Woodruff D, editors. The Chemical Physics of Solid Substances and Heterogeneous Catalysis volume 4: Fundamental Studies of Heterogeneous Catalysis. Amsterdam: Elsevier; 1982 p. 27-72 47. XPS Curve Fitting, http://www.xpsfitting.com/search/label/Platinum 48. J. Li, Z. Zhou, I. Broadwell, S. Sun, In-Situ Infrared Spectroscopic Studies of Electrochemical Energy Conversion and Storage. Accounts of Chemical Research, 2010 108 49. G. Samjeske, A. Miki, M. Osawa, Electrocaralytic Oxidation of Formaldehyde on Platinum under Galvanostatic and Potential Sweep Conditions Studied by Time-Resolved Surface- Enhanced Infrared Spectroscopy. J. Phys. Chem 2007, 111, 15074-15083 50. Sanchonx, https://commons.wikimedia.org/wiki/File:FTIR_Interferometer.png 51. M. Lefevre, E. Proiett, Iron-Based Catalysts with Improved Oxygen Reduction Activity in Polymer Electrolyte Fuel Cells. Science Vol 324 (2009) 52. F. Jaouen, J. Dodelet, Non-Noble Electrocatalysts for O 2 Reduction: How Does Heat Treatment Affect Their Activity and Structure? Part I. Model for Carbon Black Gasification by NH 3 : Parametric Calibration and Electrochemical Validation. American Chemical Society 2007, 111, 5963-5970 53. F. Jaouen, M. Lefevre, J. Dodelet, M. Cai, Heat-Treated Fe/N/C Catalysts for O 2 Electroreduction: Are Active Sites Hosted in Micropores? American Chemical Society 2006, 110, 5553-5558 54. R. W. McCabe, T. Pignet, L. D. Schmidt, Catalytic Etching of Platinum in NH 3 Oxidation. Journal of Catalysis 32, 114-126 (1974)
Abstract (if available)
Abstract
Hydrocarbon fuels contain ≈ 50 times more energy per unit mass than commercially available batteries, thus harvesting only 10% of this energy content could provide an improved power source for portable electronic devices. With this motivation, and given the ease of storage of hydrocarbons compared to hydrogen, the feasibility of using polymer electrolyte membrane fuel cells with hydrocarbon fuels, operating at low temperatures (<100˚C), was explored. The membrane-electrode assembly consisted of Nafion® N-117 for the electrolyte and platinum black powder as anode and cathode catalyst. With extremely pure (>99.99%) propane no power was produced, however, with the addition of trace quantities of unsaturated hydrocarbons, power production ensued and continued even after the unsaturated hydrocarbon addition was discontinued. Furthermore, the current history was found to have a significant influence on the cell performance. In particular, at higher current densities (> 24 mA/cm²) the power output gradually decreases then the cell rapidly “extinguishes,” however, by periodically shutting off the current for short time intervals, the average power density increased significantly. A simple model considering the relative rates of conversion of active anode catalyst sites to inactive sites and vice versa was developed to interpret this “extinguishment” behavior. This model was in good qualitative agreement with experimental data
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Design, dynamics, and control of miniature catalytic combustion engines and direct propane PEM fuel cells
PDF
Development of polymer electrolyte membranes for fuel cell applications
PDF
Development of polystyrene sulfonic acid-polyvinylidene fluoride (PSSA-PVDF) blends for direct methanol fuel cells (DMFCS)
PDF
Electrocatalysts for direct liquid-feed fuel cells and advanced electrolytes for lithium-ion batteries
PDF
Studies on the flame dynamics and kinetics of alcohols and liquid hydrocarbon fuels
PDF
Mesoscale SOFC-based power generator system: modeling and experiments
PDF
Studies of transport phenomena in hydrotalcite membranes, and their use in direct methanol fuel cells
PDF
Applications of advanced electrochemical techniques in the study of microbial fuel cells and corrosion protection by polymer coatings
PDF
Design and modification of electrocatalysts for use in fuel cells and CO₂ reduction
PDF
Investigations of fuel effects on turbulent premixed jet flames
PDF
Flame ignition studies of conventional and alternative jet fuels and surrogate components
PDF
Design of nanomaterials for electrochemical applications in fuel cells and beyond
PDF
Hydrogen fluoride addition reactions and a fluorinated cathode catalyst support material for fuel cells
PDF
Studies on direct oxidation formic acid fuel cells: advantages, limitations and potential
PDF
Accuracy and feasibility of combustion studies under engine relevant conditions
PDF
The modification of catalysts and their supports for use in various fuel cells
PDF
Studies on direct methanol, formic acid and related fuel cells in conjunction with the electrochemical reduction of carbon dioxide
PDF
Studies of Swiss-roll combustors for incineration and reforming applications
PDF
Methanol synthesis in a membrane reactor
PDF
Pressure effects on C₁-C₂ hydrocarbon laminar flames
Asset Metadata
Creator
Kong, Eugene Hyun Je
(author)
Core Title
Dynamics of direct hydrocarbon polymer electrolyte membrane fuel cells
School
Viterbi School of Engineering
Degree
Doctor of Philosophy
Degree Program
Mechanical Engineering
Publication Date
07/25/2019
Defense Date
04/26/2019
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
electrocatalysis,extinguishment,fuel cell,hydrocarbons,OAI-PMH Harvest,polymer electrolyte membrane,propane
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Ronney, Paul D. (
committee chair
), Egolfopoulos, Fokion N. (
committee member
), Prakash, Surya (
committee member
)
Creator Email
hyunjaek@hotmail.com,hyunjaekong.1990@gmail.com
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c89-191351
Unique identifier
UC11663051
Identifier
etd-KongEugene-7614.pdf (filename),usctheses-c89-191351 (legacy record id)
Legacy Identifier
etd-KongEugene-7614.pdf
Dmrecord
191351
Document Type
Dissertation
Format
application/pdf (imt)
Rights
Kong, Eugene Hyun Je
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
electrocatalysis
extinguishment
fuel cell
hydrocarbons
polymer electrolyte membrane
propane