A recent study by Joan Ogden at Princeton University’s Center for Energy and Environmental Studies tabulated a range of published estimates for the manufacturing costs of mass-produced auto fuel cell systems. The banded-structure membrane fuel cell, from Fraunhofer ISE, draws off voltage at the endplates through an integrated series connection that ties together individual cells. The typical proton exchange membrane (PEM), fuel cell consists of a series of stacked individual cells, with each cell composed of a flow field plate and a membrane electrode assembly. An air breather PEM fuel cell from DCH Technology distributes hydrogen to the flow fields through a central sleeve, while air at atmospheric pressure comes in through the sides. One of the benefits of Manhattan Scientifics’ circuit board approach is the ability of an electrode to act as a ‘preferential molecular filter’. Fraunhofer Institute scientists have been at work on a banded-structure membrane fuel cell. This device holds five-unit cells on a single plate connected as a series circuit.


Despite the strides made recently in reducing the cost of fuel cells—for instance, by decreasing the amount of platinum needed as a catalyst—they remain more costly per kilowatt than the internal combustion engine. Of course, the manufacture of fuel cells has been, and is, the domain of laboratory scientists, who assemble them by hand at fairly high hourly wages.

Until fuel cells are actually mass manufactured, their production-scale pricing can only be estimated. A recent study by Joan Ogden at Princeton University’s Center for Energy and Environmental Studies tabulated a range of published estimates for the manufacturing costs of mass-produced auto fuel cell systems. Estimates ranged from as little as $32 per kilowatt to as much as $200 per kilowatt. For the equivalent system in a gasoline-powered internal combustion engine, the cost per kilowatt was estimated at $39. Selling fuel cells to the car-buying public may prove a formidable task, at least if the case for buying is built solely upon economics.

Some companies are aiming fuel cells at the portable electronics world, where micro-power plants have a better chance than their large-scale brethren to compete on price, while offering durability, range, and stand-by capacity that even the latest manifestations of rechargeable batteries will struggle to match. Among the half-dozen or so companies pursuing miniature power, Manhattan Scientifics Inc. of New York, DCH Technology Inc. of Los Angeles, and the Fraunhofer Institute for Solar Energy Applications in Freiburg, Germany, are showcasing some remarkably novel approaches.

Patrick Turner, a research and development engineer for Manhattan Scientifics, said, “If you look at an automobile manufacturer talking about a 50-kilowatt fuel cell stack, they would like to be able to match the cost of a $500 engine from Honda, for instance. The cost of the engine works out to something like a penny a watt.” Cautioning that his example was not quite an apples-to-apples comparison, Turner continued, “A manufacturer of cellular telephones pricing up the latest lithium batteries is looking at a $10 battery averaging about a tenth of a watt over a day of use. The consumer will pay about $100 for the latest replacement battery pack. That works out to about $100 per watt for the manufacturer and about $1,000 for the consumer.”


But scaling down technology that has been concentrated mainly in the kilowatt range is not easy, said Robert Hock-aday, the company’s chief fuel-cell scientist. One problem is the expensive assembly required of traditional stacked bipolar fuel cells. Each cell uses a conductive, porous separator. And each cell usually requires four gas seals.

The typical proton exchange membrane, or PEM, fuel cell consists of a series of stacked individual cells, with each cell composed of a flow field plate and a membrane electrode assembly. A thin catalyzing layer of platinum separates hydrogen, on one side of the membrane (the anode), into protons and electrons. The protons flow through the membrane, while the electrons run through an electrical circuit. The protons and electrons reunite on the opposite side of the membrane (the cathode), combining with oxygen to produce water.

The surface area of a fuel cell determines its amperage; the greater the area, the higher the current. Voltage varies according to the number of individual cells making up a stack. More cells mean higher voltage. A conventional fuel cell to power a 6-volt cellular phone, said Hockaday, requires a stack of 12 half-volt cells and 48 gas seals.

According to Hockaday, the Manhattan Scientifics design dispenses with bipolar stacking, minimizing an expensive, painstaking assembly. Its fuel cells are manufactured in much the same manner as a circuit board. A thin plastic film is first tracked with nuclear particles. It is then etched to produce a uniformly distributed array of cone-shaped pores, each approximately 20 microns in diameter. Metals are sputtered over the substrate using thin-film vacuum deposition, forming a selectively permeable film over the pores. Vacuum deposition also creates the fuel cells electrodes and its conductive paths.

Such a technique, while making for a complex, integral cell, overcomes several problems of traditional PEM fuel cells. Rather than using four seals in every cell, for example, the Manhattan Scientifics design needs only two, one at the fuel inlet and another around the perimeter of the cell. Separator plates are eliminated, too, as are electrical junctions within the cell’s humid atmosphere. And the cell assembly is flexible enough to match a variety of end shapes.

The use of thin film as the electrical conductor does bring a penalty to the cell in the form of “significant” electrical resistance, Hockaday said. But the design of the fuel cell arrays seeks to reduce the energy-depleting effect of high resistance by organizing the cells in rows of parallel strips, like so many railroad ties, thereby minimizing the length of conductor routed between individual cells.

Rather than risk developing its fuel cells around hydrogen and the attendant possibility that such an energy carrier may have a difficult time overcoming the public perception of its dangers—Manhattan Scientifics has chosen a mixture of methanol and water for a hydrogen supply. Its design is a direct methanol fuel cell, dispensing with the need for a methanol reformer.

The company investigated the energy availability of methanol and rechargeable batteries to set minimum performance levels that a methanol-fueled fuel cell would have to meet. Hockaday said that in experiments using single cells the company has been able to exceed the specific energy of today’s lithium ion batteries.

One of the benefits of Manhattan Scientifics’ circuit board approach is the ability of an electrode to act as a “preferential molecular filter,” Hockaday said, which “allows us to manage the flow of products and reactants, similar to processes that take place in biological cells.” Methanol can poison a platinum catalyst. By decreasing the amount of methanol that hydrogen drags over to the cathode, engineers can lengthen a fuel cell’s life. Manhattan Scientifics’ molecular filter has reduced the methanol crossover rate by a factor of 60.


In 1998, Manhattan Scientifics powered a cellular phone with a prototype fuel cell system, which ran on a mix of methanol and water in equal volumes. The phone operated on standby and had enough excess power to charge a battery and make periodic calls. In 1999, the company has been developing prototype fuel cell arrays shaped to fit into cellular phones.

Air breather

DCH Technology sails a different tack with its air-breather fuel cell. Invented by Los Alamos National Laboratory scientist Mahlon Wilson and the U.S. Department of Energy, the air-breather PEM runs on hydrogen. Unlike the Manhattan Scientifics design, it uses a bipolar stack.

Each unit cell in the air-breather stack is about 2 inches across with a hole in the center. Cells stack onto a single tie bolt, which carries, in addition, a porous, channeled sleeve for hydrogen delivery. Every cell consists of an impermeable separator, flow fields for air and hydrogen, inside and outside seals, and a catalyzed membrane.

“With this geometry,” Wilson said, “the hydrogen supply to the unit cells is radially outward, and the air supply is from the periphery inward.” So the entire outer surface of the stack is a passageway for air. The design improves heat conduction for better thermal management. “All of the components in the stack are radially symmetrical, so part fabrication is simple and the entire system is potentially low-cost,” Wilson said.

Instead of designing its fuel cells to meet specific power requirements, DCH takes a modular approach, according to David Haberman, the company’s vice president for technology. “It’s unnecessary and illogical to scale PEMs to meet loads directly,” he said. It’s better to provide more power than a device actually needs. “A derated, modular fuel cell system fails softly,” Haberman said, an important consideration when fuel cells may be energizing devices that need time to store data or close programs before losing power. And by building cells in a finite number of configurations, DCH is better able to predict stack reliability.

Using this approach, DCH developed a 12-watt stack that weighs under a pound and a half and is about the size of a soda can. At the April 1998 National Hydrogen Association conference, DCH used two such stacks to operate a fluorescent light, CD player, and small television. Metal hydride canisters, from Energy Conversion Devices Inc. of Detroit, stored hydrogen for the cells. Converters stepped down the fuel cell voltage to that of the appliances.

Reliability is going to be an important measure for any integrator putting fuel cell power into its products, Haberman said. Thus, DCH’s passive fuel cell has no moving parts and needs no auxiliary equipment, such as fans for cooling the stack, pumps for pressurizing cathode air, or humidifiers for keeping polymer electrolyte membranes moist.

A significant advantage of this fuel cell—indeed, any fuel cell—over batteries is that energy and power sources are separate. “In the case of batteries, the electrodes are both the energy and power sources,” explained Wilson. That limits a battery’s range. The operating time of a fuel cell, however, is tied only to the amount of hydrogen or methanol that is stored.

Banded-Structure Membrane

Meanwhile, Fraunhofer Institute scientists have been at work on a banded-structure membrane fuel cell. This device holds five unit cells on a single plate connected as a series circuit. Plates are stacked five high, together with interleaved intermediate plates. As with the Manhattan Scientifics design, the increased ratio of active surface to sealing surface makes better use of available space than does the more traditional stacked plate design.

Hydrides Pack Away the H2

A Big issue in automotive fuel cells is the difficulty of storing sufficient hydrogen safely on board a car while giving it adequate range. For powering portable electronic devices, however, hydrogen storage is much less of a question.


Metal hydride canisters using lanthanum-nickel or titanium-iron alloys are available as standard commercial products for fuel cell developers seeking a means of storing hydrogen in a small space. The composition of these alloys can be tailored to maximize the plateau zone of the hydride. That is the region where 80 to 90 percent of the hydrogen that goes into the metal lattice also comes out, said David Tragna, project manager at Ergenics Inc., a builder of metal hydride storage devices in Ringwood, N.J.

An important consideration in designing hydride storage for portable devices, said Tragna, is the range of temperatures that the device might see. A cell phone user, for instance, could easily inflict temperatures upon the device from -10°C to 70°C, by calling in a tow for a snow-entombed car, or leaving a phone on the dashboard under the summertime sun. As a rule of thumb, Tragna said, the pressure of the hydrogen doubles for every 20°C temperature increase—though the pressure actually increases exponentially with heating. Manufacturers of metal hydride vessels reinforce the containers or modify the hydride structure to accommodate these pressure swings.

Another consideration is the limitation that impure hydrogen can impose on the number of times a hydride can absorb and discharge a gas. Metal hydrides actually make good “getters," a term Tragna used to describe a medium that purifies gases.

When the hydride metal is exposed to contaminants such as carbon monoxide or carbon dioxide, the impurities stick to the alloying particles. Used in the manner of a filter, a hydride eventually will absorb its fill of contaminants. So, a hydrogen source only 99 percent pure might limit a hydride to a hundred charging cycles, whereas an ultrapure source could stretch the hydride cycling count to 100,000 or more.

That brings up the question of charging. Is it practical for consumers to refill their own cartridges using an activation appliance, or would they be better off picking up fresh cartridges at the store? Any appliance must have its hydrogen lines cleaned and purged before charging a metal hydride vessel; if not, contamination of the cartridge is almost a certainty. Ergenics is exploring both scenarios.

According to Angelika Heinzel, who heads the energy technology department at Fraunhofer’s solar institute, miniaturization was not the initial quest of the project. Rather, Fraunhofer’s scientists were looking at ways to solve the problem of series connections by overlapping the electrode/membrane units. They patented conducting and nonconducting parts of the cell frame for the banded-structure membrane. And they found a way to combine an efficient dc converter with single cells or twin cells. “At first, miniaturization did not play the major role,” said Heinzel, “but now we are looking intensively at that topic.”

Fraunhofer displayed a fuel cell-powered laptop at the 1998 Hannover Fair. Using the banded-structure membrane cell and a separate hydride unit for storing hydrogen, institute personnel ran a laptop computer for more than 10 hours, at an output of 10 to 20 watts. That done, Heinzel said the institute is now looking at ways to “improve power density of the fuel cell so that various applications can use a fuel cell instead of a battery.” Valves and fans, she said, also need to shrink.

Hydrogen Infrastructure

Both Heinzel and DCH’s Haberman agree that the hydrogen infrastructure is the biggest barrier to commercializing these diminutive power plants. Solving that problem, for Heinzel, “means the exchange of empty hydride cartridges against filled ones at as many locations as possible.” She’s concerned also about possible problems transporting hydrogen cartridges aboard airplanes.

Said Haberman, “Many entrepreneurial fuel cell firms are busy trying to get their gizmo deployed. But it’s the hydrogen energy chain that must be made safe.” All the regulations now pertaining to hydrogen regard it strictly as an industrial chemical. “Now we must look at hydrogen in a new light, developing codes and standards for hydrogen as an energy carrier,” he said.

Even Hockaday, at Manhattan Scientifics, concedes that hydrogen may be the ultimate energy carrier for fuel cells. “The use of hydrogen in our fuel cells results in high efficiency and acceptable power density for most portable electronics,” he said. “But, hydrogen sources in the short term would not be acceptable on critical market areas such as airplane passenger cabins because of safety considerations.”


Indeed, the right infrastructure could bring relief to power-hungry laptop users. No longer would they have to wait for batteries to recharge. Fuel cells could reduce the environmental costs of battery disposal. Run time for a cellular phone could be days, not hours.

By helping people to get used to dealing with hydro-8en the way they are with gasoline—the miniaturization of power plants could make way for the onslaught of fuel cells for automobiles and stationary power generation that experts say is sure to come.

That may be why these three companies are not alone in the pursuit of small fuel cells. H-Power Corp. of Belleville, N.J., the first company to complete a commercial sale of PEM fuel cell systems, has developed a prototype portable power briefcase that provides 35 watts for laptops or other electronic devices. The company is looking at small direct methanol fuel cells, as well.


New Jersey-based Red Bank Research Co., a Bellcore-Motorola joint venture, has developed what inventor Christopher Dyer calls a thin-film fuel cell that draws both hydrogen and oxygen from a common supply, simplifying the system considerably. The fuel cell employs six unit cells, each interleaving a gas-permeable electrolyte less than a micron thick between two thin platinum electrodes. The proximity of the electrodes to one another allows hydrogen at the anode to oxidize, losing electrons, then migrate to the cathode where it rejoins with its electrons and oxygen.

In Israel, Médis El is working on fuel cells to power small electrical appliances. And industry giant Ballard Power Systems Inc. of Burnaby, British Columbia, has been exploring small fuel cells right alongside the development of automotive- and stationary-size power plants.

Much work remains in spite of the many advances being made in fuel cells for portable electronics. But fuel cells for laptop computers and cellular phones seem to have one thing going for them that fuel cells for automobiles and stationary power plants do not: strong consumer demand. The world of rechargeable electronics, said air-breather inventor Wilson, may be quicker to embrace fuel cells than either transportation or power generation because “current battery technologies are not entirely satisfactory for many portable applications due to a number of factors.” Low energy density, rechargeability, cycle life, stability, cost, and environmental issues all affect one kind of battery or another.

The current state of battery technology contrasts sharply with that of the tenacious internal combustion engine, which is inexpensive and successful, and has a sophisticated support structure already entrenched.