Increasing attention is being paid to the use of metallic materials as a replacement for non-porous graphite in bipolar plates (BPs) for polymer exchange membrane (PEM) fuel cells. Ideal bi-polar plate materials require good electrical conductivity levels, significant resistance to corrosion, good compressive strength, and reduced permeability as well as density. Despite metallic plates bearing most of these properties, they are constantly plagued with corrosion problems.
Bipolar plates are a fundamental component of proton exchange membrane fuel cells. They are designed to perform many functions, such as: distributing the fuel and oxidant in the stack; facilitating water management within the cell; separating the individual cells in the stack; carrying current away from the cell; facilitate heat management (Lee & Huang, 2003). Currently, the main commercial bipolar plates are made of non-porous graphite because of their chemical and thermal stability. However, the high price of the non-porous bipolar plates prevents them from being widely used.
Metallic materials are thought to be one of the most promising candidates to substitute for nonporous graphite bipolar plates because of their good mechanical stability, electrical conductivity, thermal conductivity, and recyclability. Also, they can be easily and consistently stamped to the desired shape to accommodate the flow channels (Lee & Huang, 2003). As a component in PEMFC, metal bipolar plates should have very high corrosion resistance because any metal ions generated from the corrosion process can migrate to the membrane, lower the ionic conductivity of the membrane, and thereby degrade the performance of the fuel cell stack. Moreover, any corrosion layer will lower the electrical conductivity of the bipolar plates and thus increase the potential loss of PEM fuel cells due to the high electrical resistance. At the anode and cathode, bipolar plates are at -0.1VvsSCE and 0.6VvsSCE, respectively (Lee & Huang, 2003). Therefore, metallic bipolar plate corrosion under the PEMFC environments is different from free potential corrosion.
The potential-pH diagram is a map showing conditions of solution oxidizing power (potential) and acidity for the various possible phases that are stable in an aqueous electrochemical system. The boundary lines on the diagram, which divide areas of stability for the different phases, are derived from the Nernst equation. Potential-pH diagrams have been used in many applications, including fuel cells, batteries, electroplating, and extractive metallurgy (Heli, 2004). We believe that this is the first time that potential-pH diagrams have been used to predict the corrosion of the metallic bipolar plates under the PEMFC working conditions.
Problem Statement
PEM fuel cells are of prime interest in transportation applications due to their relatively high efficiency and low pollutant emissions. Bipolar plates are very important components of a fuel cell and account for up to 80% of its weight. The functions of the bipolar plates are (Elmasry & Sallam, 2010).
- Provides electrical connection between adjacent cells in a stack
- Separates gases between the adjacent cells
- Facilitates water management within the cell
- Enables Heat Transfer
- Supports thin electrodes and membrane
- withstands clamping forces of stack assembly
The most widely used material for bipolar plates is graphite. Graphite is a weak and brittle material and so is not a suitable candidate for the job. Its low mechanical strength does not help it withstand the clamping forces and its brittleness makes it difficult and very expensive to machine channels with complex designs into it. These restrictions prevent fuel cells from going into mass production and becoming a major source of clean energy.
Literature review
Background information
Bipolar plates are key elements in the hydrogen fuel cell power stack, as they conduct current across cells, in addition to enhancing water management within the fuel cell. In the polymer electrolyte membrane (PEM) hydrogen fuel cell design, bipolar plates are fabricated in mass production and they must be made of materials with excellent manufacturability and suitable for cost-effective high volume automated production systems. Currently, graphite composites are considered the standard material for PEM bipolar plates because of their low surface contact resistance and high corrosion resistance (Barbir, 1995). Unfortunately, graphite and graphite composites are classified as brittle and permeable to gases with poor cost-effectiveness for high volume manufacturing processes relative to metals such as aluminum, stainless steel, nickel, titanium, etc. Since durability and cost represent the two main challenges hindering the fuel technology from penetrating the energy market and competing with other energy systems, considerable attention was recently given to metallic bipolar plates for their particular suitability to transportation applications. Metals enjoy higher mechanical strength, better durability to shocks and vibration, no permeability, and much superior manufacturability and cost-effectiveness when compared to carbon-based materials, namely carbon-carbon and carbon–polymer composites (Barbir, 1995). However, the main handicap of metals is the lack of ability to combat corrosion in the harsh acidic and humid environment inside the PEM fuel cell without forming oxidants, passive layers, and metal ions that cause considerable power degradation. Considerable attempts are being made using noble metals, Aluminum, and various coated materials with nitride- and carbide-based alloys to improve the corrosion resistance of the metals used without sacrificing surface contact resistance and maintaining cost-effectiveness.
Gold-coated titanium and niobium were the materials used by General Electric in the 1960s that were later replaced by graphite composites to reduce cost and weight. In recent years, due to lack of graphite durability under mechanical shocks and vibration combined with cost-effectiveness concerns of its high volume manufacturability, considerable research work is currently underway to develop metallic bipolar plates with high corrosion resistance, low surface contact resistance, and inexpensive mass production (Tawfik, Hung & Mahajan, 2007). Various types of metals and alloys are currently under testing and evaluation by researchers working in the field of PEM fuel cells to develop bipolar plates that possess the combined merits of graphite and metals. The ideal characteristics of a bipolar plate’s material are high corrosion resistance and low surface contact resistance, like graphite, and high mechanical strength, no permeability to reactant gases, and no brittleness like metals such as stainless steel, aluminum, titanium, etc. The main challenge however is that corrosion-resistant metal bipolar plates develop a passivating oxide layer on the surface that does protect the bulk metal from the progression of corrosion, but also causes an undesirable effect of a high surface contact resistance (Tawfik, Hung & Mahajan, 2007). This causes the dissipation of some electric energy into heat and a reduction in the overall efficiency of the fuel cell power stack.
Cost and durability are still the two pronounced challenges for the PEM fuel cell industry. The cost of large supplies of fuel cell materials and high-volume manufacturing processes must be reduced for PEM to reach an economic viability and allow it to penetrate the energy market and compete with other systems. The durability of the PEM fuel cell is another important parameter that must be improved to enhance the reliability of the two main components, namely bipolar plates and MEA. Further research and development efforts must be conducted to rectify the bipolar plate corrosion mechanisms as described below (Li & Sabir, 2005).
Metal bipolar plates
Metals such as stainless steel, aluminum, and titanium are considered for use in the manufacture of bipolar plates because of their excellent electrical conductivity and mechanical properties. However, they are normally faulted due to corrosive action they are subject to.
Corrosion failure by pinhole formation
Unfortunately, corrosion processes occur regardless whether the fuel cell is operating or not unless extreme measures are taken to evacuate the fuel cell stack of water. Corrosion failure mode is due to pinhole formation through the bipolar plate. The significant drawback with the metal bipolar plates is corrosion problems on the surface. Corrosion problems arise from the formation of an oxide coating from chemically reactive metals. These oxide layers are electrically insulating, usually on the order of 1012 ohm-cm, which imparts high contact resistance leading to a voltage drop in the fuel cell, which cannot be accepted. To avoid this voltage drop, the formation of such a resistive layer has to be prevented by coating the metal surface. The protective coating must be conductive and provide complete corrosion protection. The coating must also be chemically and mechanically stable in a fuel cell environment. The application of such a conductive, low-cost coating has been found to be very difficult to accomplish.
Corrosion failure by electro catalyst poisoning
Common electro-catalysts include Pt and Pt–Ru alloys that are susceptible to poisoning by adsorption. Most poisoning adsorbents include CO, sulfur, chloride, and low boiling point hydrocarbons. For metallic coatings on Al, little, if any at all, sulfur and chloride will be present eliminating their possible poisoning of catalysts. The same is also true for CO and hydrocarbons if a metallic coating on aluminum is used. It is possible that O2 on the cathode side may react with metal ions to form an oxide (e.g., Fe2O3, CuO) (Turner & Brandy, 2005). Since these oxides are not bound to any site, it is probable that they will be flushed from the electrodes by convective action of the air and water on the cathode. It is possible that the corrosion by-products may react chemically with the oxygen to form an oxide in the electrode (Turner & Brandy, 2005). This oxide may or may not leave the electrode causing potential fouling problems within the electrode due to blocked pores. It may be possible for H2 gas to reduce the metal ions to their metallic state. The metal deposits would be located in regions where the proton may reside including the liquid phase water and the ionomer coating of the catalysts. In a similar manner to the cathode, the metal deposits may flush out of the electrode via the convective flow action.
Corrosion failure by passivation formation
The overall comprehensive testing and evaluation of various materials for metallic and non-metallic bipolar plates are clearly compiled to provide a quick reference of the most up to date research findings in this area of PEM fuel cell technology.
Review of corrosion issues in metallic bipolar plates for PEMFCs
In PEMFCs applications, corrosion of metallic bipolar plates is one of the mostly researched issues because as the metallic bipolar plates corrode, metal ions are released and block the ion conduction mechanism for H+ at the membrane (Turner & Brandy, 2005). In addition, the contact resistance will increase due to the oxide layer formation at the surfaces (Tawfik, Hung & Mahajan, 2007). Finally, the pinhole type of corrosion (i.e., pitting corrosion) can further lead to accelerated corrosion due to the concentration cell formation (i.e., two different potentials occurring on the same plane) (Turner & Brandy, 2005). All these phenomena caused by corrosion of the metallic bipolar plates would directly lead to performance degradation and shortening stack life.
The United States Department of Energy (DOE) has placed some goals for metallic bipolar plates to be used in fuel cell stacks. The corrosion goals are set as follows. At the anode side the corrosion current at 0.1V vs. SHE (standard hydrogen electrode) while hydrogen is bubbling through an acid solution (representing the acidic fuel cell environment) should not exceed 1 μA/cm2 (Tawfik, Hung & Mahajan, 2007). On the other hand at the cathode side the corrosion current at 0.85V vs. SHE while oxygen is bubbling through an acid solution should not also exceed 1 μA/cm2. In addition, the goal for the cost of a metallic bipolar plate is suggested to be less than 6$/kW (5).
Coated metals
Metallic bipolar plates are coated with protective coating layers to avoid corrosion. Coatings should be conductive and adhere to the base metal without exposing the substrate to corrosive media (Li & Sabir, 2005).Two types of coatings, carbon-based and metal-based, have been investigated (Turner,H. & Brandy, 2005). Carbon-based coatings include graphite, conductive polymer, diamond-like carbon, and organic self-assembled monopolymers (Turner,H. & Brandy, 2005). Noble metals, metal nitrides and metal carbides are some of the metal-based coatings. Further, the coefficient of thermal expansion of base metal and the coating should be as close as possible to eliminate formation of micro pores and micro cracks in coatings due to unequal expansion (Turner,H. & Brandy, 2005). In addition, some coating processes are prone to pinhole defects and viable techniques for coating bipolar plates are still under development (Turner,H. & Brandy, 2005). Mehta and Cooper presented an overview of carbon-based and metallic bipolar plate coating materials. Carbon-based coatings include: graphite, conductive polymer, diamond like carbon, organic self-assembled monopolymers. Metal-based coatings include: noble metals, metal nitrides, and metal carbides (Turner,H. & Brandy, 2005).
Barbor and Gomez concluded that the coefficient of thermal expansion (CTE), corrosion resistance of coating, and micro pores and micro cracks play a vital role in protecting bipolar plates from the hostile PEM fuel cell environment. The authors also argue that even though PEM fuel cells typically operate at temperatures less than 100 ◦C, vehicle service would impose frequent start up and shut down conditions, and temperature differentials of 75–125 ◦C would be expected during typical driving conditions. A large difference in the CTE of the substrate and coating materials may lead to coating layer failure. One technique to minimize the CTE differential is to add intermediate coating layers with CTEs between that of adjacent layers (9). Materials such as Al, Cu, and Ni, are very susceptible to electrochemical corrosion in acidic solutions that are typical of PEMFC operating conditions. However, materials such as Au and phosphorous Ni show very high resistance to electrochemical corrosion, comparable
Research methodology
A number of properties that makes a material suitable for fuel cell operation will be evaluated. These include conductivity, mechanical properties, corrosion resistance, and mass of the sample and its cost of production. The materials to be evaluated will include Titanium, Aluminum, Stainless Steel, Graphite, and Moldable Polymer Composite.
Measurement of in-plane and through-plane conductivity
In-plane (bulk) conductivities will be measured according to ASTM Standard F76- 86. Current contacts will be placed at the four corners of the plaque allowing for a constant current to pass through the specimen. The voltage drop will be measured across the specimen with a Keithley 2000 digital multi-meter at ambient conditions (Wang, 2006). Two characteristic resistances, RA and RB will be measured. The plaque resistance, RS, is obtained by solving the Van der Pauw equation: exp(-πRA/RS)+exp(-πRB/RS)=1…….(1)
The resistivity, ρ, is given by p=RSd , where d is the thickness of specimen. The volume conductivity, σB, is defined as 1/ρ.
Through-plane conductivities will be measured based on a method proposed by Landis and Tucker. [20] A 76.2 mm x 76.2 mm plaque was placed between gold plated copper electrodes. Between the electrodes and sample was placed a piece of carbon paper (Toray TGP-H120) to improve electrical contact between the electrodes and sample. The system was placed under a compaction force of 2000 pounds (approximately 1000 psi) and the resistance was measured. The sample is removed, and the resistance of the test cell (including carbon paper) was measured again under the same conditions to obtain a “baseline” resistance (Wang, 2006). The sample resistance could then be calculated by subtracting the baseline resistance from the total resistance. The resistivity of the material was calculated by: p=((RT-R baseline) A)IL(2)).
Where p is resistivity, A is cross-sectional area of sample, L is the thickness of sample, and RT and R baseline are total resistance and baseline resistance, respectively. The through-plane conductivity, σT, was then calculated as 1/ρ.
Measurement of half-cell resistance
To measure the half-cell resistance of a bipolar plate, an apparatus will be set up similarly to the one used to measure through-plane conductivity. A bipolar plate having dimensions 12.2 x 14.0 x 0.3 cm and an active area of 100 cm2 will be placed between the gold plated copper electrodes. Carbon paper (Toray TGP-H-120) will be placed on both sides of the bipolar plate, and hence, in between the sample and electrodes. The size of the carbon paper will be 10 x 10 cm in order to completely cover the active area. The sample will be placed under a compaction force of 2000 pounds (approximately 130 psi) while a constant current of 250 mA will be passed through the sample (Wang, 2006). The potential will be measured between the collectors and the half-cell resistance will be calculated based on Ohm’s law. The bipolar plate will be removed, and the potential across the electrodes and carbon paper will be measured to produce a baseline resistance. The baseline resistance that is the resistance of the testing circuit excluding the bipolar plate but including carbon papers and electrodes will be measured after testing of the plate. This will be done to ensure the stability of the baseline of the instrument and to evaluate the contribution of the bipolar plate to the total half-cell resistance.
Mechanical properties of materials
In addition to electrical conductivity, the bipolar plates should also have adequate mechanical properties and resistance to creep to be used in fuel cell stacks where they would be subjected to a constant compressive load. However, it is difficult to obtain high conductivity and sufficient mechanical properties simultaneously. As a result the mechanical properties of most materials used to produce bipolar plates are still lower than the target values.
Corrosion Test Methods
Corrosion potential of a substrate can be obtained from the open circuit potential of the metal substrate in liquid electrolyte (Wang, 2006). In order to observe the behavior of the electrode in varying potentials, potentiodynamic experiment should be conducted. This experiment shows us if the electrode is in active or passive state for corrosion (Wang, 2006). This experiment also provides initial corrosion current values at certain potentials regardless of time (Barbir, 1995). Corrosion current values at certain potentials can be obtained with this method very quickly. A linear sweep voltametry experiment is made between -0.5V and
1V at 1mV/s sweep rate. All potentials mentioned here are compared with respect to the standard hydrogen electrode. At a certain potential, the graph would give a peak, which represents the corrosion potential value, while the corrosion current could be obtained from the tangent curve of the negative and positive over-potential curves (Wang, 2006).
In this study, both potentiodynamic and potentiostatic experiments will be carried out for stamped, and hydro-formed bipolar plates in a custom made corrosion cell controlled with a PAR 2200 potentiostat. To resemble the PEMFC environment, the metallic bipolar plates will be placed in an acidic liquid electrolyte tank which contains 0.5 M H2SO4 solution at 80°C. This corrosion cell consists of a specimen holder where a metal blank [bipolar plate] is held inside and acts as a working electrode, Poco graphite counter electrode, an Ag/AgCl reference electrode, and a gas bubbler at the bottom. All of this setup is placed inside a temperature controlled environment.
In order to remove the residual oil and dirt from the surface, all specimens will be cleaned with acetone before testing. Sensitive cleaning will be also applied to remove the oil in channel grooves by placing the specimens in an ultrasonic bath filled with acetone for 30 minutes. The non-active area and the back part of the test specimens will be covered with Teflon tape, and thus, only the active area of the bipolar plate will be exposed to the acid. The working electrode [i.e., bipolar plate] will be placed in the corrosion cell before placing the electrolyte. All connections inside the cell will be covered with Teflon tape to prevent the corrosion from the sulfuric acid vapor. Additional 0.5 M H2SO4 will be also placed inside the temperature controlled environment with a closed tap to prevent excessive evaporation of the acid at 80°C. A potentiodynamic experiment with no gas, a potentiodynamic experiment with one of the gasses and a potentiostatic experiment with the same gas will be carried out continuously for each test sample. The acid level will be controlled several times during the experiment to submerge the working electrode under the electrolyte throughout the test period. The potentiodynamic experiments will be performed by increasing the potential from – 0.5V to 1.0V at a rate of 1mV/s (Barbir, 1995). All of the potentials are given vs. Standard Hydrogen Electrode [SHE]. Three different potentiodynamic conditions will be carried out which include: purge no gas [acidic condition], purge H2 [anode side condition], and purge O2 [cathode side condition]. On the other hand, the potentiostatic experiments represent the steady state operation condition of the fuel cell and the current value will be obtained and recorded when the system reaches steady state while keeping the potential constant at 0.1V vs. SHE during H2 purge and at 0.8V vs. SHE during O2 purge
Analysis of results
Analysis will focus on evaluation of the materials that present the best characteristics necessary for functionality of bi-polar plates. Among the key characteristics for evaluation will include conductivity, permeability, and cost of production, corrosion resistance and mechanical strength. For portability mass will also be considered.
References
Barbir, K. (1995). “Progress in PEM fuel cell system development.” In: Yurum Y. (Ed), Hydrogen energy system, NATO ASI Ser. E., Vol. 295, pp. 203–213.
Chaudhuri, T. & Spagnol, P. (2005). Bipolar plates for PEM fuel cells: A review. International Journal of Hydrogen Energy, 30, 1297-1302.
Elmasry, M. & Sallam, M. T. (2010). Structure and Mechanical Properties of Aluminum Metal Matrix Composite Produced by Hot Pressing Technique.SAT-13-MS-14, 2, 2010, 23-45.
Heli W. (2004). Turner: Ferritic Stainless Steels for Bipolar Plate for Polymer Electrolyte Membrane Fuel Cells, Journal of Power Sources 128, 193-200.
Lee, S. J. & Huang, Y.P. (2003). Metal Processes: Coating aluminum to enhance properties. Processes Techno, 140, 2003.
Li, X. & Sabir, I. (2005). Review of bipolar plates in PEM fuel cells: Flowfield designs. International Journal of Hydrogen Energy, 30, 359-371.
Tawfik, H. Hung, Y. & Mahajan, D. (2007). Metal bipolar plates for PEM fuel cell – a review. Journal of Power Sources, 163, pp.755–767.
Turner,H. & Brandy, M. P. (2005). Corrosion Protection of Metallic Bipolar Plates for Fuel Cells, May 22–26, DOE Hydrogen Program Review,
Wang, D. O. (2006). Northwood. An Electrochemical Investigation of Potential Metallic Bipolar Plate Materials for PEM Fuel Cells, Presented at International Symposium on Solar-Hydrogen-Fuel Cells, Cancun, Mexico.