Mighty Mites

Leroy Ohlsen stared at a TV screen, engrossed. The college freshman was captivated not by the Superbowl or a Hollywood thriller, but rather by a Discovery Channel documentary showing a bus cruising city streets, powered by hydrogen fuel cells and emitting nothing worse than water vapor.

The experience was a watershed in his life. While he pursued a bachelor's degree in chemistry at the University of Washington (Seattle), he "dove into all the journals and read everything I possibly could on fuel cells—to understand them, especially the catalysis aspect, where platinum and ruthenium do the magic of busting up chemicals to generate electricity."

Some day before too long, we may all be glad he did. Ohlsen is now in the vanguard of a small group of innovators who are reinventing the fuel cell, the battery-like device in which hydrogen from a fuel reacts with oxygen to produce water and electricity.

Over the past few years, Ohlsen and his competitors have logged several breakthroughs in fuel cells powered directly by alcohol and alcohol mixtures. Though these devices are unlikely to power buses anytime soon, they may well be the key to liberating small appliances from the tyranny of rechargeable batteries. Imagine, for example, a laptop that operates all day or for the better part of a trans-Pacific flight instead of conking out before the meal service dishes are cleared away. Even better, imagine recharging (or, rather, refueling) the device in just a few seconds instead of the hours required to recharge a battery.

Ohlsen's company, Neah Power Systems Inc. (Bothell, Wash.), is all of four years old. He started it in partnership with an old high-school buddy, Michael Fabian, right after earning his B.S. in 1998, to try to build a better fuel cell. Today, his company's experimental porous-silicon, methanol-fed fuel cells seem to have some important advantages over more traditional designs based on other technologies, like polymer membranes. Neah claims its approach will result in much higher power density—and therefore higher energy density, because space not taken up by the fuel cell stack itself can be used to house a bigger fuel tank. But those other designs, too, are constantly improving.

Batteries have troubles...

The market Ohlsen plans to compete in—replacements for batteries in handheld equipment like cellphones, PDAs, and laptop computers—is exciting for three main reasons:

• Handheld devices are demanding more power than ever. As manufacturers keep adding features like large color screens and wireless interconnectivity, power consumption goes up despite heroic efforts at power management.
• Although batteries are better than ever, they are not keeping up with the demands of handheld equipment. Even the latest lithium-ion batteries can power a laptop computer for only a few hours.
• More important, batteries take hours to recharge and require an electrical outlet to do it. They're not a good solution to the problem of keeping a digital camera up and running during a two-week trek up the Amazon.

Alcohol-fed fuel cells, referred to as DMFCs—for direct methanol fuel cells, although they sometimes run on other kinds of alcohol or alcohol mixtures—are more practical than hydrogen cells for portable applications. The reason: they have no need for high-pressure hydrogen tanks or other special fueling equipment. As their name implies, DMFCs use alcohol directly as a fuel. They do not first extract its hydrogen in a separate piece of equipment and then use the gas.

Perhaps their biggest advantage over batteries is that they are quick and easy to refuel. Most manufacturers plan to offer two refueling options: replace one sealed cartridge with another or open a cap and squirt fuel in, much like refilling a cigarette lighter. Either way, the process is much less burdensome than plugging something into an electrical outlet for several hours.

...but then, fuel cells have problems, too

But fuel cells are not without problems. Their big technical problem is improving the system energy density—the number of watt-hours they store per liter of overall volume—especially when operated at environmental temperatures much above or below normal room temperature.

For the most part, technical progress is occurring at an encouraging pace. According to Atakan Ozbek, director of energy research at Allied Business Intelligence Inc. (Oyster Bay, N.Y.), a technology market research and consulting firm, fuel cells' most serious issues center more on government regulation and marketing than on technology.

The very concept of passengers' being allowed to carry concentrated methanol on aircraft requires approval from regulatory agencies like the U.S. Department of Transportation and the U.S. Federal Aviation Administration. At present, concentrations greater than 24 percent are banned from the passenger compartment. Efforts are under way by fuel cell companies to have the regulations modified, a process that may take anywhere from six months to two years, according to Ozbek. Approval will doubtless come, as it did for lithium-ion batteries in the early 1990s, but until it does, no one expects the market to begin to take off in any serious way.

Even when it does start to move, in 2005, Ozbek expects only some 25 000 units to be produced worldwide. But later on, watch out! He expects production to reach 3 million units in 2008, 50 million in 2010, and 200 million in 2011.

Marketing is the other big issue, figuring out exactly who will be willing to pay extra for what long-running device. Before a large consumer market develops, fuel cell manufacturers will probably concentrate on niche products like high-end wireless bar code scanners and field service instruments.

Three main approaches

In making DMFCs today, manufacturers are competing by using three main technical approaches. The best known, which is being pursued by large companies like Motorola, Samsung, and Toshiba, is based on a thin sheet of plastic known as a proton exchange membrane (PEM). It gets its name from the fact that it allows hydrogen ions (protons) to pass through the membrane, but bars electrons. The heart of a typical PEM fuel cell consists of three layered parts:

• A negative electrode, or anode, to which fuel is fed and which is impregnated with a catalyst that breaks the fuel down into carbon dioxide, hydrogen ions, and electrons.
• A polymer membrane that allows the passage of protons but blocks the flow of electrons.
• A positive electrode, or cathode, to which air is fed and which is impregnated with a catalyst that promotes the combination of oxygen, protons, and electrons into water vapor.

The overall effect of what goes on in such a fuel cell is that the fuel is oxidized. But instead of the oxidation occurring in one place and in one grand reaction, the polymer membrane forces the electrons to get to their desired destinations via an external circuit where they can be harnessed to do useful work. So the reaction is broken into two half-reactions, one on each side of the solid polymer membrane electrolyte.

PEM-based fuel cells have historically been plagued by many problems. Water management has perhaps been the biggest, requiring several pumps and valves to move some of the wastewater generated at the cathode back to the anode. There it can be used to dilute the incoming fuel, since too high a concentration of alcohol can cause crossover—diffusion of unburned fuel across the membrane to the cathode, where it would be consumed without generating any electric power.

Orientation of the cell has also been a problem, since the collected water could migrate where it wasn't wanted under the influence of gravity. Today, thanks to clever (but confidential) architecture and the choice of particular structural materials, companies like Mechanical Technology Inc.'s subsidiary, MicroFuel Cell Inc. (MTI/MFC, Albany, N.Y.), are building PEM-based DMFCs with just a single valve or pump.

In a laboratory demonstration, MFC recently operated a fuel cell using undiluted "neat" methanol, thereby maximizing its energy density. The simple system used no pumps or valves for water management.

Further along the development cycle is the MFC prototype shown on the opening page, which has a total volume of 80 cubic centimeters, comprising a planar array of six series-connected cells, a fuel tank, and electrical control and conversion circuitry. It puts out around 11.5 Wh from one little tank of fuel, for a volumetric energy density of about 145 watt-hours per liter. This is not yet as good as a lithium-ion battery, but the figures are improving fast and CEO William Acker expects to reach parity later this year.

Neither membrane nor methanol

An alternative approach using a liquid electrolyte, much like an ordinary battery, has been taken by Medis Technologies Ltd., whose technology originated in the Soviet Union. Now headquartered in New York City, the company is developing its technology in a laboratory in Israel.

The Medis design has several special aspects. It uses neither a membrane nor methanol, but substitutes for them a liquid electrolyte and a blend of ethanol and other materials. According to CEO Robert K. Lifton, doing so allows Medis to completely avoid the problem of methanol toxicity experienced with PEM technology. The Medis design also requires no control systems for feeding fuel, temperature control, or water management.

The cell uses very little precious metal as a catalyst—a tiny bit of platinum on the negative electrode and none on the positive. Lifton believes that reducing the amount of noble metals compared with PEM technology gives his company's technology a crucial advantage in the price-sensitive consumer market.

One unusual feature of Medis's products is that they are single fuel cells. The company does not make stacks or planar arrays of low-voltage cells to get higher voltages. Instead it takes one large (high-current) cell and boosts its 0.5 V to whatever is wanted in a dc-dc converter. By eliminating stacks, it avoids a lot of heat transfer problems. Specifically, it has no need for forced-air cooling.

Lifton hopes to follow a razor-and-blade model in his business: sell the fuel cell itself as cheaply as possible and then make money on the (intrinsically very cheap, but patented) replacement fuel cartridges. He also believes that he will have less trouble with regulatory agencies with a nonmethanol fuel cell than with a methanol-based one.

Meanwhile, back in Washington, Leroy Ohlsen has come up with an approach that combines porous-silicon electrodes and a liquid electrolyte. Each silicon electrode is riddled with millions of pores. Since the interior surface of each pore is coated with a catalyst, the Neah structure brings a very large catalytic surface to bear. This means that fuel crossover isn't a problem because the distributed catalyst will handle all the fuel to which it's exposed.

The structure, although very fine, is not terribly expensive because it is made using standard semiconductor fabrication techniques. (Intel Corp., Santa Clara, Calif., is an investor in the company.) The two separate electrodes are bonded together using a dielectric material so they do not form a short circuit. Aside from the dielectric around the edges, the space between the electrodes is filled with a liquid electrolyte, which is sealed in.

Methanol is fed to the outer surface of one electrode while oxygen is applied to the other. As the methanol passes through the pores of the negative electrode, it gives up its carbon dioxide, sends its electrons out to an external circuit, and passes its protons across the liquid electrolyte. The protons enter the positive electrode, where they meet up with the electrons (via the external circuit) and an oxygen atom in the presence of a catalyst. The result: water.

Importantly, the Neah fuel cell does not get its oxygen from the air, but rather from hydrogen peroxide, making possible a fully sealed fuel system in which all of the input materials—methanol and hydrogen peroxide—are contained in the fuel cartridge (in separate compartments, of course), and all of the waste products—mainly water and carbon dioxide—are collected in that cartridge for disposal. The benefits of this arrangement include improved reliability with no exposure to potential external contaminants, and no water exhaust, which can be a problem with other designs if water vapor is vented too close to some electronic components, according to Neah.

Because handheld equipment is demanding increasing amounts of energy—more than batteries of acceptable size, weight, and cost can provide—the future for micro fuel cells looks pretty bright. Rechargeable batteries are far from dead, but the DMFC is presenting them with a challenge that cannot be ignored.