Chapter 8 Materials for Proton Exchange Membrane Fuel

Chapter 8 Materials for Proton Exchange Membrane Fuel Cells (PEMFCs)

1. Introduction Proton exchange membrane (PEM) fuel cell technology is a promising alternative for a secure and clean energy source in portable, stationary, and automotive applications. However, it has to compete with cost, reliability, and energy efficiency with established energy source such as batteries and internal combustion engines. Many of the major challenges in PEM fuel cell commercialization are closely related to three critical materials considerations: cost, durability, and performance. The challenge is to find a combination of materials that will give an acceptable result for the three criteria combined.

State-of-the-art PEM fuel cells, using thinner membranes (< 40 μm) and Pt/C electrodes (< 1 mg Pt/cm 2) for cost reduction, are less expensive (but still higher than DOE cost targets) but only have a demonstrated lifetime of less than 15, 000 h operating on reformate. The idea of using an ion-conductive polymeric membrane as a gaselectron barrier in a fuel cell was first conceived by William T. Grubb, Jr. (General Electric Company) in 1955. In his classic patent, Grubb described the use of Amberplex C-1, a cation exchange polymer membrane from Rohm and Haas, to build a prototype H 2 -air PEM fuel cell (known in those days as a solid-polymer electrolyte fuel cell).

Today, the most widely used membrane electrolyte is Du. Pont’s Nafion due to its good chemical and mechanical stability in the challenging PEM fuel cell environment. A perfluorinated polymer with pendant sulfonated side chains, Nafion was initially developed in 1968 by Walther G. Grot of Du. Pont for the chlor-alkali cell project of the National Aeronautics and Space Administration (NASA) space program. Several manufacturers provide other perfluorinated polymers, composite polymers, and hydrocarbon polymers as membrane electrolytes.

A membrane electrode assembly (MEA) usually refers to a five-layer structure that includes an anode gas diffusion layer (GDL), an anode electrode layer, a membrane electrolyte, a cathode electrode layer, and a cathode GDL. Most recently, several MEA manufacturers started to include a set of membrane subgaskets as a part of the MEA package. This is often referred to as a seven-layer MEA. In addition to acting as a gas and electron barrier, a membrane electrolyte transports protons (H+) from the anode, where H 2 is oxidized to produce H+ ions and electrons, to the cathode, where H+ ions and electrons recombine with O 2 to produce H 2 O.

For economic reasons, air is usually used as the cathode feed rather than pure O 2. Electrons are carried from the anode to the cathode through the external electric circuit. The anode and cathode electrode layers are typically made of Pt or Pt alloys dispersed on a carbon support for maximum catalyst utilization. Ionomers and polytetrafluoroethylene (PTFE) resins can be added to the electrode layers. The former extends the proton transport path beyond the electrode-membrane interfaces；latter facilitates liquid water removal from the electrode layers. Both can also help bind together various electrode components.

GDLs are made of porous media such as carbon paper or carbon cloth to facilitate the transport of gaseous reactants to the electrode layers, as well as the transport of electrons and water away from the electrode layers. An MEA is sandwiched between two bipolar plates to form a single fuel cell. The word bipolar refers to a plate’s bipolar nature in a series of single cells (known as a stack) in which a plate (or a set of half plates) is anodic on one side and cathodic on the other side. Bipolar (half) plates often have gas channels on the side facing an MEA and channels for temperature control on the other side and, together with the GDLs, they provide structural support for MEAs in addition to serving as transport media for reactants/products, electricity, and heat.

2. Electrode Materials The hydrogen oxidation reaction (HOR) occurs at the anode electrode of an H 2/O 2 PEM fuel cell (reaction 1): H 2 2 H + + 2 e - E 0= 0 V This is a thermodynamically reversible process that often serves as a standard reference electrode, known as the reversible hydrogen electrode (RHE), for all other electrochemical process.

At the cathode electrode, thermodynamically irreversible fourelectron oxygen reduction reaction (ORR) is the dominant electrochemical process (reaction 2): + - O 2 + 4 H + 4 e 2 H 2 O E 0 = 1. 229 V When connect through an external circuit, the net result of these two half-cell reactions is the production of H 2 O and electricity for H 2 and O 2. Heat is also generated in the process. In the absence of a proper catalyst, however, neither of these two half reactions takes place at meaningful rates under PEM fuel cell operating conditions (50 to 80 °C, 1 to 5 atm). Despite decades of effort in search of cheaper alternatives, platinum is still the catalyst of choice for both the HOR and ORR.

Anode Catalyst Materials As mentioned above, an anode serves as the HOR site in a H 2/O 2 fuel cell. As such, it must fulfill the following basic functional requirements: 1. transport H 2 to the catalyst sites 2. catalyze the HOR process 3. carry proton away from the reaction sites to the membrane electrolyte 4. remove electrons from the anode 5. transfer heat in or out of the reaction zone

The HOR process is believe to proceed through the following steps (Reactions 4 to 6: M = metal catalyst): H 2 + 2 M → 2 MHads Tafel reaction or + H 2 O + M → MHads + H 3 O + e and + - MHads + H 2 O → H 3 O + e + M (4) - Heyrovsky reaction Volmer reaction (5) (6) The rate-determining step varies depending on the specific catalysts and the reaction conditions. For a PEM fuel cell with a Pt anode, the HOR process involves only the Tafel and Volmer reactions, with the Tafel reaction being the rate -determining step. The rate of the overall HOR process can be express in the Bulter-Volmer form (equation 12. 1): j = j 0 [ e (1 -β 1) F ηs /RT – e β 1 F ηs /RT ] (12. 1)

The exchange currant density (j 0) depends on the natural of the catalyst morphology, the catalyst-electrolyte interface, the properties of the reaction media (p. H, electrolyte, temperature, concentration, etc. ), and the levels of contaminants such as CO, Cl-, and sulfur species. For the HOR, the measurement of j 0 is further complicated by the lower H 2 gas diffusivity in a strong electrolyte solution known as the “salting out” effect. As a result, the reported value ranges from 10 -5 to 10 -2 A/cm 2 for different Pt electrodes in various acidic media. However, the ratelimiting process of a H 2/O 2 fuel cell is the ORR on the cathode because j 0 for the ORR is 10 -6 ~10 -11 A/cm 2. Anode materials research has been centered mostly on Pt-loading reduction, CO-tolerant catalysts for DMFCs and systems operating with CO-contaminated fuels (such as reformate), and low-cost Pt alternatives.

Cathode Catalyst Materials A cathode serves as the site for the ORR in a H 2/O 2 fuel cell. It should fulfill the following basic functional requirements: 1. transport O 2 to the catalyst sites 2. catalyze the ORR process 3. carry proton from the membrane electrolyte to the catalyst sites 4. remove electrons to the reaction sites 5. remove product water 6. transfer heat to or from the reaction zone

The exact ORR mechanism is still a topic of much debate. Two representative mechanisms, namely, a dissociative mechanism (reactions 11, 15 and 16) and an associative mechanism (reactions 12 to 16), are illustrated here： The final two steps are the same for both mechanisms:

Recent studies pointed to the formation of a peroxy intermediate on the Pt surface, suggesting that the more complex associative mechanism is at work on a Pt electrode. However, the DFT calculations by Nørskov at al. suggested that the associative mechanism was only the dominant pathway at ORR overpotentials greater than 0. 8 V. At realistic ORR overpotentials (< 0. 8 V), the two pathways run parallel to each other. Regardless of the actual ORR mechanism, one thing is clear: the ORR is the rate-limiting process in a H 2/O 2 fuel cell.

The slow ORR kinetics is responsible for the steep slope in the activation polarization region (Figure 12. 2). A small but noticeable H 2 crossover current at the open circuit (OC) is largely to blame for the opencircuit voltage (OCV) loss even for a state-of-the-art membrane electrolyte. Much of the cathode research has been directed at finding a cathode catalyst to improve the slow ORR kinetics and to find a cheap replacement for Pt. Pt and Pt alloys are the most active catalysts for the ORR.

Stability of Pt Cathode Catalysts Cathode lifetime durability presents a special challenge in PEM fuel cells. The major problems associated with the catalysts are Pt agglomeration, sintering, dissolution, and redistribution. The cathode environment is highly oxidative and corrosive due to high voltage (e. g. , 0. 6 to 1. 0 V), low p. H, elevated temperature, and the presence of water and oxygen. Although Pt has low solubility at normal cell operating voltages, its solubility increases significantly with cell voltage and reaches the highest dissolution rate at round 1. 1 V.

Furthermore, freshly formed Pt. Ox and Pt(OH)x are less stable in acidic media than material that have been aged. This makes a PEM fuel cell very susceptible to Pt dissolution when its voltage cycles are between 0. 75 and 1. 2 V. Some dissolved Pt ions (e. g. , Pt 2+) will migrate into the ionomers or membrane. Where they are reduce to Pt by H 2 that has diffused from the anode side. These Pt particles in membrane/ionomers are unlikely to participate in the ORR process because of the lack of electrical continuity. Some dissolved Pt can redeposit into other Pt particles, which results in the growth of the Pt particles. It is believed that the exact band location is determined by the H 2 crossover rate.

Electrode Support Materials The primary functions of a good catalyst support are to (1) maximize catalyst utilization, (2) transport electrons, and (3) transfer heat. Other desirable attributes include high chemical and electrochemical stability, good mechanical integrity, and, last but not least, low cost. Carbon materials, such as Vulcan-X 72 by Cabot Corp. , are widely used as PEM fuel cell catalyst supports because of their high surface area, low cost, excellent electric/thermal conductivity, and adequate stability and mechanical properties. The use of carbon-supported Pt in place of Pt black is directly responsible for the reduction of Pt loading in PEM fuel cells.

However, this also substantially increase involve gaseous reactants (H 2/O 2), protons, and electrons, the active catalyst sites must have access to these species at the same time. Such reaction sites are often referred to as catalyst-electrolyte-reactant triple-phase boundaries (TPBs). In order to increase catalyst utilization and reduce the activation overpotential loss, it is critical to maximize the TPB regions within an electrode. Proton-conducting materials (such as Nafion ionomers) are usually added to the electrode layer to improve the proton transport beyond the electrode-membrane interface, thus increasing the Pt utilization. The ionomers and PTFE resin also serves as a binder to keep the Pt/C particles together and to create hydrophobic/hydrophilic domains for facile gas and water transport. Others have attempted to use more stable carbon materials such as graphitized carbon and carbon nanotubes (CNTs).

3. Membrane Electrolyte Materials Membranes are a critical and challenging component in PEM fuel cell applications. In order to meet the cost and durability target for residential and automotive applications, a membrane electrolyte must meet several functional requirements: 1. high proton conductivity over a wide RH range 2. low electrical conductivity 3. low gas permeability, particularly for H 2 and O 2 (to minimize H 2/O 2 crossover) 4. good mechanical properties under cycles of humidity and temperature 5. stable chemical properties 6. quick start-up capability even at subzero temperatures 7. low cost

The life time of a membrane is related to its original thickness, its mechanical integrity, and the chemical stability of the constituent ionomers. There are many factors that can affect the membrane life time. These include not only the membrane properties (such as its chemical structure, composition, morphology, configuration, and fabrication process), but also several external factors (such as contamination of membrane, materials compatibility, and the operating conditions of the fuel cell).

Perfluorosulfonic Acid Membrane Materials Perfluorinated membranes are still regarded as the best in the class for PEM fuel cell applications. These materials are commercially available in various forms from companies such as Du. Pont, Asahi Glass, Asahi chemical, 3 M, Gore, and Solvay. Perfluorosulfonic acid (PFSA) polymers all consist of a perfluorocarbon backbone that has side chains terminated with sulfonated groups.

Polybenzimidazole (PBI) Membrane Materials Plug Power, working with its European partners PEMEAS (now BASF Fuel Cell Gmb. H) and Vaillant, is actively pursuing a PBIbased HT PEM fuel cell system as a CHP system with high system efficiency and great CO tolerance. PBI (see chemical structure above) is a hydrocarbon membrane that has been commercially available for decades. Free PBI has a very low proton conductivity (~10 -7 S/cm) and is not suitable for PEM fuel cell applications. However, the proton conductivity can be greatly improved by doping PBI with acids such as phosphoric, sulfuric, nitric, hydrochloric, and perchloric acids.

4. Gas Diffusion Layer Materials The GDL functional requirements can be summarized as follows: (1) allow uniform transport of reactant gases to the electrode； (2) conduct electrons；(3) remove product water from the electrode；(4) transfer heat to maintain the cell temperature； and (5) provide mechanical support for the MEA. To fulfill these functions, an ideal GDL material should have small gas transfer resistance, good electron conductivity, and good thermal conductivity. The porosity of a GDL structure is the most important parameter for reactant transfer. Water management is the most challenging problem in GDL and fuel cell design. The ability to remove water is one of the key properties in evacuating GDL performance.

The cell performance can be improved by placing a fine layer (such as a microporous layer (MPL)) between the GDL and the catalyst layer because it creates a saturation jump across the interface. Commonly used GDL materials are made of porous carbon fibers, including carbon cloth and carbon paper. Carbon cloth is more porous and less tortuous than carbon paper and has a rougher surface. Experimental results showed that carbon cloth GDL has better performance under high-humidity conditions because its low tortuosity (of the pore structure) and rough textural surface facilitate droplet detachment. However, under dry conditions, carbon paper GDL has shown better performance than carbon cloth GDL because it is capable of retaining the membrane hydration level with reduce ohmic loss.

5. Bipolar Plate Materials Bipolar plates play an important role in fuel cell operation. Generally, the functions of bipolar plates can be summarized as (1) supply and separate reactant gas without introducing impurities； (2) conduct electrons； (3) remove the heat and control the fuel cell operating temperatures； and (4) remove the product water from the system. To fulfill these functions, an idea bipolar plate material should have minimal gas permeability, high electrical conductivity, high thermal conductivity, and good chemical and electrochemical stability in the fuel cell environment. From a manufacturing perspective, the bipolar plate material should have low density, good mechanical strength, easy processability, and low cost. Nowadays, materials commonly used for bipolar plates are either graphite based or metallic.

Up until the early 1990 s, pure graphite was used as the primary bipolar plate material for its good electron conductivity, low density, and high corrosion resistance (e. g. , 680 S/cm and 1. 78 g/cm 3 for POCO AXF-5 Q). However due to its poor strength and ductility, graphite plates cannot be easily fabricated, and the minimum thickness is limited to about 5 mm. This result in a large stack volume and low efficiency in heat transfer. The brittleness of the graphite limits the methods of fabrication, which lead to a high manufacturing cost.

6. Materials Compatibility and Manufacturing Variables

References Materials for the Hydrogen Economy, Jones, R. H. and Thomas, G. J. , ed. , CRC Press, Boca Raton, 2008.