Developing optimal fuel cells by Takehiko Yaza, Seika Machinery Executive overviewIn developing new fuel cells, there are many obstacles to overcome. For example, gas diffusion must be improved, the volume of catalysts must be reduced and recondensation must be prevented. Another major difficulty is managing water in the fuel cell so that the proton exchange membrane is moisturized and water is quickly discharged from the cathode. Failure to control the water discharge properly will cause flooding at the cathode and a reduction of voltage. There will be a considerable difference depending on the materials selected and the treatment and processing of the materials. Failure to understand the properties of each individual material will make it impossible to develop the optimal fuel cell (Figure 1). February 25, 2010 – Key factors to consider when developing new fuel cells include: controlling the gas diffusion, permeation, condensation, the reactive area, as well as moisturizing and discharging water. All of these factors depend on the particle size, fiber diameter of the materials and the through-pore structures in them at the electrodes. This article introduces tools that enable better understanding of these properties. Figure 1. Current issues for developing fuel cells. Milling In North America, ball mills are the most common method for reducing the size of catalyst particles and micro-porous layers of slurry. In a ball mill, the pulverizing energy is generated by steel balls in a gravity-based process. Generally, the milling energy is 1G. By contrast, in a bead mill, the dispersing slurry contains carbon blacks, catalysts and beads, which are mixed together by an agitator inside the milling vessel where particle communion breaks them down into smaller particles. The high-speed of the stirring device generates 100-500G of centrifugal force, providing very high energy. The bead mill makes a large difference in the velocity between the beads and the milled particles. Occasionally, the beads and slurry rotate together and the slurry is not pulverized efficiently. To solve this problem, the AIMEX Alpha Mill (Figure 2) uses an orifice contractile flow vessel. In this configuration, even low-speed rotation generates a very large velocity difference between beads and the slurry. By applying this principle, the milling operation can be carried out very effectively using 30μm beads. Consequently, micrometer-sized carbon black and platinum catalyst particles can be transformed into particles with nm sizes. Figure 2. a) Alpha Mill generates high ΔV with small power; b) comparison with the existing model. When evaluating the efficiency of the slurry pulverization by particle size distribution, we recommend using Horiba Partica LA-950V2. This instrument has the closest correlation to data for NIST traceable to particle size standards and it provides the widest measuring range in the industry. When the average diameter of the particle to be pulverized is ~0.5μm, even if the materials are crushed, agglomeration is prone to happen. To prevent this phenomenon, a dispersant can be added to the slurry, but this method does have limitations. Therefore, maintaining a high degree of dispersion while coating catalyst and micro-porous layers becomes more important and the Ultrasonic System PRISM series meets this requirement. This coating system uses ultrasonic energy and a nozzleless feed system to eliminate nozzle-blocking problems. The result of these modifications is a 1mm uniform coating thickness. In screen printers and other typical coating systems, it is difficult to provide a thin and uniform thickness without using a pressing machine. Understanding pore structure It is necessary to fully understand the pore structure of the coated catalyst layer and the micro porous layer, and the properties relative to water. Although the traditional method of mercury porosimetory generally is used to measure the opened pore structure and volume, the high-pressure measurement method will destroy the pore structure and cannot pick up the functional pores on the application. In contrast, the PMI Capillary Flow Porometer series can measure pore size distribution to determine gas diffusion, water permeability, and the repellent characteristics of the gas diffusion and catalyst layers, without the use of mercury and liquid nitrogen. It realizes automatic and short duration measurements based on the ASTM, bubble point and half-dry methods. Figure 3 compares measurements made by the Mercury Porosimeter and the Capillary Flow Porometer on a polymer membrane that has only through-pores. The former shows 1-4μm of pore size distribution (peak pore size: 3μm) while the latter shows 1-1.2μm peak pore size:1.1μm difference. Figure 3. a) Measurement pore size distribution by Mercury Porosimeter; b) Measurement pore size distribution by Capillary Flow Porometer. Gas diffusion Specifically, gas diffusion is the phenomenon in which high-concentration gas flows towards low concentration gas, and eventually the concentration becomes even in the differentiated space with a concentration gradient. Differential pressure results in permeation, but this cannot be called diffusion. There are several measuring instruments for the water vapor transmission rate in the market, however, these instruments can measure only membranes without pores. When fuel cells use high permeability samples, it is difficult to evenly control the pressure at the primary and secondary sides, and the measurement instruments that can measure diffusion at high accuracy are not available. The Seika Moisture Vapor Diffusion Permeameter (MVDP) is integrated using the technologies to maintain high humidity to some degree and prevent condensation by means of temperature control and pressure control. Instead of using batch methods in gas chromatography, MVDP uses real-time technologies to measure concentrations. As shown in Figure 4, the equipment can measure water vapor diffusion of gas diffusion layers and a proton exchange membrane, as well as oxygen diffusion of moisturized gas. In addition, it can measure gas permeability of generated condensation in the inside of a sample by simulating the flooding phenomenon. Figure 4. a) Measurement in-plane gas permeability test for carbon papers by MVDP; b) oxygen diffusion test for carbon papers by MVDP. Many fuel cell materials are treated with a waterproofing chemical on the surface to prevent condensation, or are hydrophilically processed to retain moisture. The characteristics of Teflon-coated samples change with immersion time and temperature. The Seika Liquid Intrusion Meter measures the changes in properties and helps optimize fuel cell performance (Figure 5). A liquid permeability test using ultra-low differential pressure can monitor permeated water from large through pores in order of pore size. The Mesys USM-200 thickness gauge is effective for the management of thickness for the proton exchange membrane manufacturing process. This method emits no radiation, and does not come in contact with, or destroy the samples during measurement. Figure 5. a) Hydrophobic sample measurement; b) Hydrophilic sample measurement. Conclusion Because it is difficult to visualize the inner operation of a fuel cell — it is covered by separators made of metal or carbon — it is necessary to use measurements (to get structural information) and simulations (to get process information). Using these direct approaches brings in new ideas and compensates simulation results on super computers. Furthermore, using the aforementioned instruments, we developed simulators for monitoring gas permeability in chronological order on blocking pores by condensation (flooding) at GDL, while also creating a gas visualization system for gas diffusion and the concentration at GDL. However, these are only one set of requirements that need to be addressed before high-volume manufacturing of high-efficiency, cost-effective fuel cells can be realized. Going forward, we would like to discuss additional research proposals with interested parties and propose new equipment necessary to develop optimal fuel cells. Acknowledgment Teflon is a registered trademark of DuPont. Biography Takehiko Yaza has a bachelor of arts in economics from Takasaki City U. of Economics. He is a senior sales manager at Seika Machinery Inc., 3528 Torrance Blvd., Suite 100, Torrance, CA 90503 USA; ph.: (+1) 310.540.7310; e-mail firstname.lastname@example.org; www.seikausa.com.