meeting report
The Small Fuel Cell Conference
New Orleans, LA USA
Dennis Sieminski
- Portable Energy Think Tank
Smyrna, GA
The Small Fuel Cell Conference sponsored by The Knowledge Foundation (www.knowledgefoundation.com, Brookline, MA, phone: 617-232-7400) was held in New Orleans on May 7-9. This conference has been an annual event since 1999. The attendees are representatives of companies, universities and government agencies from a number of countries whose focus is commercializing small fuel cells for portable electronic devices. This year’s attendance was about 200. A worthwhile opportunity in the exhibit area is to see, first hand, several operational prototype small fuel cells. This synopsis of the conference is arranged into five sections: the Nanomaterial Workshop, discussions on the small fuel cell market, companies commercializing fuel cell products, R&D at universities, and government agency work.
Nanomaterial Workshop
A pre-conference workshop occupied the first half-day of the conference and four presentations dealt with the subject of carbon nanostucture materials. Carbon nanomaterials are a class of polymers of pure carbon in new spatial configurations. The configuration that occupied much of the workshop was the Single Wall Nanotube (SWNT). This material can be produced as an extremely high aspect ratio tube (e.g., diameter of ~1nm, length of 1k to 10k nm) where the wall thickness can be one carbon atom with the surface of the tube a tile structure of hexagonal ring carbon molecules.
Carbon nanomaterials are an exciting development in material science because the properties they offer are extraordinary – strength 100x steel, conductivity similar to copper, surface area >2000m2/g, and thermal stability 500EC in air, 1400EC anaerobic. Applications for carbon nanomaterials include conductive inks and adhesives, high performance composites, flat panel displays and fuel cells.
Within the fuel cell category, there are multiple applications for carbon nanomaterials – anode and cathode structures with high conductivity and very finely dispersed Pt; high weight percent (4-6%) hydrogen storage medium; bipolar plates that are light, tough, highly conductive, easy to fabricate; and corrosion-resistant interconnects. The presentations dealt primarily with SWNTs as a catalyst support structure for fuel cell anodes and cathodes.
Carbon Nanotechnolgies Inc. of Houston, Texas, has a depth of intellectual property in the area of dispersion of catalysts in nanomaterials. David Karohl, director of business development, offered some very encouraging test results on PEM fuel cell anodes made with C nanomaterial that suggested excellent performance could be achieved with Pt loading levels about 1/100th of current anode materials. More work needs to be done to validate these initial results, and more compelling testing of the material in cathodes is planned.
Dr. Thomas Gennett, professor of chemistry, reviewed the Rochester Institute of Technology’s (RIT) capabilities for making SWNTs. A large variety are possible (e.g., diameter, length, level of impurities, chirality are some defining characteristics). Central to the successful development of nanomaterials is working through the maze of variables in the raw materials, processing methods and settings to arrive at material with the desired set of properties for the intended application. RIT is establishing methodologies for making and characterizing nanomaterials. One easy to communicate example of this is the SEM and TEM pictures, which reveal SWNT shapes and catalyst particle dispersion. RIT closes the development loop by working with commercial partners and government agencies that test the materials in different fuel cell applications. Here some performance gains for nanomaterials versus conventional materials are starting to be seen.
Dr. John Erkey, associate professor of chemistry at the University of Connecticut, is investigating carbon aerogels as a cathode material for PEMFC. These materials are mesoporous, with high surface area, good electrical conductivity and can be produced in a variety of shapes. Initial results with cathodes were promising: 0.5mg Pt/cm2 delivered 0.4V@1000mA/cm2.
Dr. John R. Regalbuto, associate professor of chemical engineering at the University of Illinois at Chicago, spoke on the fundamentals of Pt catalyst impregnation for carbon materials, introducing the revised physical adsorption (RPA) model which is an electrostatic mechanism and model where surface charge, point of zero charge (PZC) and proton balance are key factors in describing the absorbing anionic and cationic forms of Pt. Pt anions have high uptakes at low pH and Pt cation uptake is high at high pH. To control metal adsorption properties by altering the PZC of a substrate, they first tried ion doping of silica and alumina, but redissolution of the dopant foils this method. So, the next attempt will look at different oxidations of C surfaces.
Meeting between sessions are (from left): Shimshon Gottesfeld, CTO of MTI Micro Fuel Cells; Ashish Pattekar, Lehigh University doctoral candidate; and David Karohl of Carbon Nanotechnologies.
The Market for Small Fuel Cells
Discussions on the market for small fuel cells covers issues such as: where they will be applied, how can their advantages be leveraged and how do they compete with batteries in portable devices. This will be an ongoing forum until prototypes and beta product can actually test the market. Presentations by several fuel cell companies later in the program provide an indication of where their thinking is in this process and where they stand in terms of technical achievement.
Mark Hamden-Smith of Superior MicroPowders and Kurt Kelty of Panasonic Technologies dealt with the question of where fuel cells will replace the incumbent power sources – batteries and internal combustion engine (ICE) generators. Kelty’s presentation very solidly makes the case for continued dominance by batteries in small portable devices on the basis of cost, consumer habits, regulatory barriers, and technical feasibility. He allows that large form factor applications that can bear a price premium is the sweet spot for fuel cells. Hamden-Smith makes an important point on the form factor and size point, which can often be easily missed. He illustrates how a direct comparison of battery to fuel cell energy density is not always a sufficient exercise because typically this does not adequately address applications where multiple batteries must be carried to meet the runtime requirement. Here the advantage of fuel cells becomes pronounced. Early market entry points are identified as military, grid unavailable applications (e.g., emergency services, camping), APU (auxiliary power units), scooters and E-bikes, which are driven by emissions concerns.
Companies Developing Small Fuel Cells
Dr. Steven Harford, senior systems engineer at Ball Aerospace & Technologies Corp., reported that the company has been working on fuel cells for military applications for ten years. In that time, the business has grown from $0.5 to $15 million; this includes hydrogen fuel cells, DMFC, SOFC, fuel source development and testing. Military fuel cell requirements are more severe than the typical consumer applications. These devices must survive rugged shock, vibration, temperature, orientation, and water immersion tests. Their main competition is primary batteries in high watt-hour configurations, e.g., three-day missions requiring 1500Wh. By comparison, a portable computer battery is typically much less than 100Wh. In military applications, fuel cells look very attractive because the high watt-hour requirement makes them smaller and lighter than batteries. Plus, the focus is more on refueling cost rather than initial cost. Ball has in the field 50W and 100W direct hydrogen PEM FCs that are 50% of the weight of the complement of LiSO2 primary batteries that would be needed. Their next design direction is lower power 20W DMFCs with a form factor of 6x4x1.75”, weight of 0.6kg not including fuel, and is soldier wearable (objective force warrior, OBW). These will be 20% the weight of equivalent primary batteries needed for the job. Methanol cartridges are being developed in 0.25 liter/0.25kg/400Wh and 0.50 liter/0.6kg/800Wh sizes. Ball’s technical expertise in cans and packaging will be leveraged here.
The fuel cells in a presentation by Frank Ignazzitto, V.P. of sales and marketing at Avista Labs, represent the high side of what might be considered small fuel cells and are really more in the stationary category. However, this does serve as a good reference point for a number of reasons. Avista is a fuel cell business with ongoing sales and product in the field, demonstrating the technology and market equation needed for a viable business. The applications served is power for when the grid goes off, or there is no grid – telecom, UPS, standby, emergency power. A typical application needs 10 to 100 hours per year of backup. The impact of power failure ranges from the costly to the catastrophic. Backup batteries and internal combustion engine (ICE) generators serve the present market. A main disadvantage for fuel cells is their initial capital outlay – much higher compared to the battery option; however, they can have a lower 10-year life cycle for certain specified runtimes. The challenge of the business is that fuel cells must be able to deliver at today’s prices. The “soft” advantages of size, clean, quiet are not sufficient to win sales. Furthermore, the business cannot be overly reliant on volume to drive costs down. The product engineering is quite elegant and must be seen to be appreciated with modular, hot swappable cartridges providing self-humidification and superior reliability. Systems are typically sized for power, and hydrogen bottles are added to meet runtime.
Motorola Energy Technology Labs is working on two systems in a parallel effort – DMFC and a reformed methanol FC. A reformer has certain advantages as power requirements go up. Working on both systems enables Motorola to understand the tradeoffs. Plus, each approach would likely have different applications. Jerry Hallmark, manager of the technology lab, talked about the micro methanol reformer project. Allison Fisher, principal staff scientist, talked about DMFC stack performance.
Tel Aviv University researcher Arnon Blum (left) meets FCT author Dennis Sieminski of Portable Energy Think Tank.
Demonstrating his company’s pump is William Donaldson, president of Par Technologies.
The micro fuel processor converts liquid methanol (MeOH) to hydrogen and will feed an elevated temperature PEM FC. Multi-layer ceramic technology is used for fabrication. They don’t use Si MEMs since there is fast turn around with ceramic and it definitely is inert to everything. The unit is thermally integrated with the FC and insulated. The fuel is two parts MeOH, one part water. There is no CO cleanup. They are able to control the reforming reaction to keep CO below 2% and everything is run back through the combustor. Steam reforming takes place at 230EC, with hydrogen gas delivered at 150EC at 10-25 micro-liters/minute. Dimensions are 35x15x4mm and up to 55x55x4mm to supply a four-cell stack that delivers 250mA @ 2.4V; this is 200mA/cm2 @ 0.6V/cell, ~130 mW/cm2. At 150mW/cm2, the target is 800-1000Wh/liter.
Motorola has tested its DMFC four-cell planar Nafion MEA configuration to determine what is required for flow fields, stoichiometric flow rates, single pass and fuel recirculation. The system uses pure MeOH, mixing it with water from a collecting system. The conclusions on stoichiometry are: at the cooling anode S~7 higher increases crossover and voltage degradation; lower leads to CO2 blinding which starves the last cell in the series. At the cathode S~5 higher results in cooling and low V; lower results in flooding and low V. The operating architecture is critical for life. They have 1000 hours on their most recent 1W system.
Manfred Stefner, CEO of Smart Fuel Cells, reported that the company is doing a market test that bears watching. They are introducing two standardized DMFC product offerings. One is small enough to fit in a briefcase to power a mobile office. The other is luggable size and can be used for activities like camping. The devices run on 100% MeOH. In their offering, Smart Fuel Cells has addressed several important market issues. They have standardized a product, completely integrated it, obtained CE and TUV certification, and provided a fuel refill source for their customers. They have designed a special cartridge for MeOH and obtained IATA certification for air shipment, so their customers can receive direct delivery of fuel from them. The units are designed for the consumer market with an attractive appearance and simple operation. Specs for the two units are provided in a table below along with expected user applications. The fuel cells will compete with batteries, ICE generators and photovoltaics, which are today’s established products in these applications.
Dr. Hyuck Chang of Samsung’s Advanced Research Material and Devices Lab in Korea, provided an interesting perspective on small fuel cells since Samsung is not only a fuel cell company, but a major consumer electronic products company. Driving the company’s fuel cell program are considerations of product direction in portable computers and cell phones. In portable computers, Samsung sees a bifurcation in product size over time driven by function. One path will have an emphasis on mobility and as a consequence devices will continue to get smaller, thinner and more mobile, culminating in some type of wearable device. The other product path emphasizes high performance (e.g., more storage, screen size, speeds) and long runtimes. The form factor here is expected to get larger over time, pushing out today’s laptop-size envelope. In cellphones, projections of energy requirements out to 2010 for 3G and 4G devices show a linear increase over time with more space for the battery needed to maintain acceptable runtimes.
Samsung’s fuel cell development road map starts with portable computer products being developed first in 2004-06 with 10-50W, 300Wh/l, 2000-hour life. They are followed by 4G cellphone fuel cells in 2006-2009 with 2-5W, 500Wh/l, 5000-hour life.
Performance benchmarks for current prototypes under development are: a 5W DMFC /battery hybrid for 4G cellphones measuring 95x70x29mm using a MeOH/water mixture, and an air mover 174ml/min, fuel pump 0.24ml/min. Cells are in a planar arrangement. For portable power, they have 30, 70 and 100W stacks. The MEA is a rigid hybrid membrane with includes a low MeOH crossover membrane (<10% Nafion), a CO-tolerant MeOH oxidation catalyst, a MeOH tolerant oxygen reduction catalyst and a high surface area high conductive carbon nanomaterial. Power density is 75mW/cm2 @ 30C, 120mW/cm2 @ 50C, 2M MeOH, run at 3x stoichiometric air.
Robert Hockaday, chief scientist at Energy Related Devices and Manhattan Scientifics, explained that passive operation is highly desirable in a fuel cell because it has such a strong positive influence on many aspects of design complexity, size, weight, and cost. So if passive operation is the chosen design pathway, then diffusion principles must be a major tool. Diffusion domains include fuel and oxygen delivery, discharge product removal, electrolyte and crossover blocking. Manhattan has demonstrated some diffusion fuel cell proof of concepts. A MeOH (99.8%) ampoule fed, 4-cell, 1mW device on which they have logged 511 days of continuous operation with 8 ampoule changes, which give a cumulative 180Wh/kg of fuel. A one-cell system shows 400Wh/kg of fuel. Several other conceptual models are being explored – a cylindrical fuel cell (looks like a D cell battery) with a fuel tank core wrapped by an air manifold surface; and a 1W cellphone charger. One product close to completion is a small heater that can deliver 1W/cm2 with diffusion-controlled automatic thermostat.
Shimshon Gottesfeld, CTO of MTI MicroFuel Cells, reported that the company is targeting 2004 for introduction of its first commercial fuel cell product. They presently have two prototypes that run on 50% MeOH/water mixture. The objective is to get to neat MeOH operation.
At the conference, the company demonstrated its prototype hybrid fuel cell/Li-ion battery running a Nokia 3650 cellphone/camera. The unit has 10Wh of energy and delivers 0.5W with a peak of 3W. Refueling at this point is accomplished by injecting fuel into the tank. A second larger prototype being developed for Harris Corp. to run a military radio has an output of 5W with 25W peak. Energy is 50Wh. MTI’s goal is to exceed the energy density of primary BA 5590 batteries.
DMFC power presently stands about 100mW/cc of stack, fuel is 50% MeOH/water, fuel utilization is 90%, cell efficiency is 27%, oxygen is obtained by passive exposure of the cathode to ambient air, a planar stack arrangement consists of five cells. In laboratory testing, they have been able to demonstrate completely passive systems that use neat MeOH with >80% fuel utilization resulting in ~900Wh/l of fuel.
Medis Technologies is noteworthy since it is one of the few companies going down the path of developing an alkaline fuel cell system. Gennadi Finkelshtain, general manager, discussed advantages of the company’s system such as the ability to avoid high loading with precious metal catalysts, good power density, and less comlex water management. Plus, Medis is using ethanol as a fuel, which does not have the safety issues associated with methanol. The main obstacle of the alkaline system has been carbonation of the alkaline electrolyte by atmospheric CO2, which deteriorates power and, by extension, life. Medis has dealt with this problem in part by making electrolyte refreshes coincident with refueling, and developing a system to operate passively with no pumps and no forced air.
Medis has developed a fuel cell for General Dynamics’ ruggedized military PDA that operates with the rechargeable PDA battery. A prototype was demonstrated at the conference. The 125x80x40mm pack delivers 2W, 5V@0.4A. The use pattern for the power pack is a 72-hour mission with refueling by a cartridge required every 12 hours, i.e., six cartridges required. The energy in a cartridge is 14Wh. Plans call for a boost in cartridge energy so that changes will be less than four.
Technical directors meet: Brian Walsh (left) of Breakthrough Technologies and Robert Wichert of U.S. Fuel Cell Council.
Medis/More Energy’s Mark Estrin, engineering manager, and Gennadi Finkelshtain, general manager, take booth duty.
University Development Programs
Tokyo Metropolitan University’s methanol crossover development program was discussed by Dr. Kiyoshi Kanamura, professor of applied chemistry. Methanol crossover, in which a percentage of unreacted MeOH is able to pass from the anode to the cathode through Nafion membranes, is a major technical problem being investigated. One solution is a methanol barrier, which can be created by screen-printing a layer of polybenzimidazole (PBI) to Nafion. Another approach is modifying the surface structure of Nafion by using electron beam (EB) irradiation. A third strategy involves suppressing the expansion of the polymer electrolyte membrane as it absorbs MeOH. An inorganic-organic composite membrane provides a ceramic matrix with connected uniform three-dimensionally ordered pores in the polymer electrolyte membrane to suppress the expansion of the gel polymer and reduce MeOH crossover. An organic gel electrolyte (AMPS) is an aqueous mixture of 2-Acrylamido-2-methylpropane-sulfonic acid, N,N-methylenebisacrylamide and ammonium peroxodi sulofate.
To generate hydrogen for a 20W fuel cell using MeOH requires about 2W of heat for the endothermic reaction. Mayuresh Kothare and Ashish Pattekar of the Integrated Microchemical Systems Lab at Lehigh University reported that the university has developed a micro reactor with catalyst particles packed in serpentine microchannels with integrated temperature sensors and heaters. The device includes a Pd micro membrane separator for the water-gas shift reaction. It has attained 90% conversion of MeOH. Hydrogen delivery is sufficient for an 8W fuel cell. The external setup requires precise pumping of the reactant liquids; a mass spectrometer is used for product gas analysis. Experimental results are used to further improve analytical math models. Insulation is an important design consideration since significant losses through convection occur at the high temperature surfaces involved.
James Fenton, professor of chemical engineering at the University of Connecticut, presented information on the university’s high temperature MEA development program. Nafion, the current state-of-the-art fuel cell membrane, develops significantly higher resistance as operating temperatures go above 90EC. The cause is lower ionic conductivity as water loss occurs. High temperature membranes attempt to overcome this problem; plus, they promise a number of other advantages – higher performance because of accelerated kinetics at higher temperatures, better CO tolerance which reduces requirements on fuel purity, and easier water management. To get to this point, dehydration of the membrane and dry out of the catalyst layer at higher temperature operation (>100C) must be overcome. The main path is developing composite membranes by adding proton-conducting salts to Nafion. Examples are: Nafion- Teflon- phosphotungstic acid (NTPA); Nafion-Teflon-zirconium hydrogen phosphate (NTZP); Nafion-zirconium hydrogen phosphate (NZP). In addition to the high temperature membrane work, they are also investigating improvements in the catalyst layer and enhanced gas transfer in the hydrophobic/hydrophilic regions of the gas diffusion layer. Present fabrication is one at a time in glove boxes with plans for scaling up.
In his discussion regarding the University of Puerto Rico’s development program for GDLs for high W/high Wh/l DMFCs, Dr. Eugene Smotkin, professor of chemistry, explained that DMFC performance is temperature-dependent. Two main contributing factors are Arrhenius dependence at the anode and higher crossover at low temperatures. Of primary concern is that optimal power density and energy density (MeOH conversion percentage) operating points occur at different locations as functions of temperature, MeOH flow rate, MeOH concentration. The presentation uses 3D graphs to illustrate this point. There is fairly good coincidence at high temperature, but optimums for power and energy occur at opposite conditions at low temperature. The conclusion is that addressing the anode is required; specifically, alleviating mass transport in the carbon fabric gas diffusion layer and improving the limiting current by a redesign of the current collector.
Government Agency Development Programs
MTI Micro Fuel Cell’s George Relan compares notes with Greg Fritz of Energy Conversion Devices.
Dr. Christel Roux revealed that the Small Power Sources Laboratory of the French Atomic Energy Commission (Commissariat a l’Energie Atomique, CEA) is engaged in fundamental research activity and prototype development of a MEMs fuel cell with the aim of providing technological transfer to an industrial partner. The main objective of the program is to develop commercial power sources for portable equipment that would provide significant advantages over Li-ion batteries. The approach is to manufacture fuel cells using microelectronic fab processes to avoid the high cost, size and weight associated with conventional fuel cell construction. They are developing an integrated stack on 4-inch Si wafers. The presentation illustrated the elements of the fab process and results of the first prototypes > 50mW/cm2, using air, hydrogen fuel, RT. The program is looking at both hydrogen and MeOH as fuels. In addition, they are testing a solid alkaline fuel cell (SAFC), compatible with heavy alcohols (MeOH, ethanol, ethylene, glycol, glucose); it uses a membrane, which eliminates carbonation effects.
Canada National Research Council’s development program for micro fluidics was the focus of a presentation by Dr. Kevin Stanley, microcell project leader at the Institute for Fuel Cell Innovation. Stanley explained that there is a significant body of work for modeling macro fuel cell fluid behavior, but fundamental changes must be considered for the hydraulic diameters found in small fuel cells, e.g., pressure-velocity relations, mixing, bubble formation and behavior. Typical dimensions for fluid flow in small fuel cells are on the order of 30-500Fm and important Dimensionless Numbers have values like Reynolds Number<1, Knudsen Number<.001. Effects to be aware of are: larger pressure penalties since pressure drop is a function of the square of the diameter, boundary layer effects become much more important, most mixing occurs through diffusion since laminar versus turbulent flow is typical; for gases this is okay because complete mixing occurs in 1.5mm for 300 Fm channel but liquids require centimeters of length.
Dr. Christopher Hebling of the Fraunhofer Institute in Germany discussed the university’s objective to develop the scientific, engineering and production foundation for miniature, passive fuel cells. They have done studies, performed experiments, developed math models and built prototypes. Their experience covers a wide range of fuel cell technical issues, e.g., flow field geometries, current distribution, flooding, and catalyst efficiency. They also have investigated various production technologies for components, e.g., ultrasonic and photochemical machining, micro milling, micro embossing, micro injection molding.
ISSUE MAIN >