Micro fuel cells have been a promising technology for years. Are they ready to start delivering on their promise?

Click here to enlarge image

Micro fuel cells offer a potential replacement for batteries in portable applications. Invented in 1839, fuel cells first gained public attention when NASA used them in the 1960s for the Apollo space missions. Fuel cells have been powering submarines since the 1980s and buses and cars since the 1990s. The question is whether their little brothers, micro fuel cells, should be taken seriously in portable applications any time soon.

What is a fuel cell?

First off, a fuel cell is not a battery. A battery is an energy storage device that provides power as needed. It “starts up” instantly, needs lengthy recharging, and wears out after 300-500 charge cycles. A fuel cell, in contrast, is a power generator that provides constant power, needs time to warm up, can be recharged quickly, and wears out after several thousand hours.

Figure 1. Operation of a direct-methanol fuel cell. (Figure courtesy of Toshiba)

A fuel cell burns hydrogen in the presence of oxygen, produces electricity, and generates water, CO2, and heat as byproducts. There are three main types of fuel cells:

•Direct hydrogen, which obtains hydrogen fuel from compressed gas, metal hydride, or chemical hydride

•Fuel processing, which externally processes hydrocarbon fuels that are rich in hydrogen, such as methanol (reformed methanol fuel cell, or RMFC) to generate hydrogen on demand, which is then fed into the cell

•Direct conversion, which directly uses liquid hydrocarbons, such as concentrated or neat methanol (direct methanol fuel cell, or DMFC), from which the hydrogen is separated in the cell itself.

All fuel cells use the same basic structure: an anode and a cathode separated by a liquid or solid electrolyte. A catalyst, such as platinum, is often used to speed up the reaction at the electrodes.

There are two major system architectures for portable fuel cell systems: active and passive. Active systems have pumps to push air onto the cathode; condensers to separate out the water vapor that comes off the cathode; tanks to mix it with incoming methanol to maintain the proper mixture; and pumps that move the resulting fuel to the anode. Passive systems eliminate all of these components and rely on fuel injected directly onto the anode and ambient air circulating to the cathode. Passive systems are less complex and expensive than active ones, but they generate less power and are less efficient.

The type of fuel cell most suited to small portable applications is the DMFC. These cells (Figure 1) can directly use liquid methane at the anode, where the methanol is first split into hydrogen and carbon dioxide before a special platinum/ruthenium catalyst splits the hydrogen into protons and electrons. The protons then diffuse across the polymer membrane to the cathode, while the electrons pass as current through the external circuit. At the cathode the electrons recombine with the protons and with oxygen to form water, which, along with the carbon dioxide, is then passed out of the system. The voltage generated by a single DMFC cell is 0.3-0.9 V, though cells are typically arranged in stacks (Photo 1) of varying sizes, depending on the application requirements.

Photo 1. A fuel cell stack.

DMFCs have a potential advantage over LiIon batteries that has proven elusive: higher energy density. While pure methanol has a 10X greater energy density than LiIon by volume and 30X by weight, in practice this advantage remains theoretical. DMFCs typically operate at about 30% efficiency, wasting as much as 70% of the energy produced as heat. Much of the problem is due to the tendency of methanol to cross over the membrane into the cell instead of reacting at the anode.

To counteract this problem, many manufacturers dilute the methanol, but this in turn reduces the energy density of the cell. This can result in larger cells to achieve adequate power output, further diminishing the attractiveness of fuel cells versus batteries.

To compare energy densities in a meaningful way, you need to plot weight against runtime. If the LiIon battery pack in your notebook computer will power it for four hours, how much will the fuel cell and the required fuel weigh that can power it for the same duration-and how large will they be? With current technologies, your notebook battery will show a higher energy density than a fuel cell. But if you increase the timeframe-forcing you to add more and more batteries to complete the “mission”-the fuel cell’s density quickly passes that of the batteries, since the weight of the additional fuel is far less than the weight of additional batteries. This may seem an odd scenario when discussing a laptop computer, but it is quite meaningful if you’re a foot soldier carrying a large load of batteries for an extended period-a key target market for fuel-cell manufacturers.

Fuel-cell makers take different approaches to dealing with the limitations of DMFCs. Using a patented circulation system, MTI Micro Fuel Cells’ Mobion DMFC runs on pure methanol, eliminating water as an emission. MTI claims that its Mobion test units, customized to the same size and shape as the U.S. Army’s standard BA5590 portable battery, ran continuously on one integrated fuel tank to achieve twice the energy output.

Neah Power Systems makes a silicon-based DMFC, replacing the 10-mil polymer membrane with a 400-micron silicon honeycomb containing a catalyst, giving a much larger reactive area. Neah uses an enclosed, all-liquid system, using methanol as the fuel and nitric acid as the oxidizer. With an eye toward the military market, Neah claims its cells can decrease the weight soldiers have to carry related to portable power by up to 65% over a 72-hour mission.

Photo 2. Toshiba’s DMFC-enabled MP3 player.

UltraCell Corp.’s UltraCell25 attaches hot-swappable cartridges to a DMFC that runs on 65 % methanol. The UltraCell device places a reformer stage that takes in methanol and outputs hydrogen to the cell stack. Ultracell claims that its cell with two refill cartridges achieves an energy density of 340 Wh/kg versus LiIon’s 160 Wh/kg, yielding a 50% weight saving (3.5 lb versus 7.3 lb) when running a 20-W laptop for 24 hours.

Not surprisingly, manufacturers of portable electronic devices are looking seriously at micro fuel cells. Motorola, Toshiba, NEC, Fujitsu, Hitachi, Samsung, Sanyo, and LG, among others, all have active fuel cell programs, have demonstrated prototypes, and are considering the commercial viability of their products. At this point Toshiba has gone farther than most and recently demonstrated a DMFC-powered portable audio player (Photo 2) that runs for four hours on a charge, at which point you recharge it with a squirt of pure methanol from a pocket dispenser. The company has also demonstrated a DMFC-powered cell phone, which it is developing with Japanese carrier KDDI. The phone weighs 150 gm, of which the fuel cell contributes 40 gm. Toshiba claims the cell generates 7 Wh, enabling 2.5X longer talk time than would be possible with a LiIon battery.

While technically viable, micro fuel cells won’t start stealing major market share from LiIon batteries in cell phones, PDAs, and notebook computers anytime soon. Part of the reason is the nature of fuel cells: they’re power generators, not batteries. They put out constant power, and most portable devices have variable power requirements. If your fuel cell is putting out 20 W when your notebook only needs 5 W, you’re wasting power.

It’s possible to meter the fuel just as a car meters gas to the engine, and manufacturers are doing that using the SMBus in notebooks. But it’s not yet possible to respond to power surges as quickly as a battery, so you’d need to have the fuel cell put out more power than you need at any given time-unless you use the fuel cell in combination with a battery, which may make the most sense in many applications, such as recharging your notebook battery on a flight from New York to Beijing.

DMFCs produce water vapor as a byproduct-not something you want to introduce in an electronic device. They can also generate a fair amount of heat, another no-no. Active or reformer-type cells, while generating higher power than passive DMFCs, have noisy pumps and fans. DMFCs use exotic metals (usually platinum) as a catalyst, making them relatively expensive. And durability remains an issue in both active and passive cells. Finally, while the fuel-cell engine may be comparable in size to a battery, the fuel adds bulk, weight, and cost. Now that rechargeable batteries have all but replaced disposable ones in portable devices, will consumers be willing to pay to run their devices?

There are regulatory issues, too. No U.S. or European airline will allow passengers to carry methanol cartridges onboard. In October 1995 the U.S. Fuel Cell Council offered proposed standards to the International Civil Aviation Organization (ICAO), which they hope to see approved in 2006.

Many of the objections raised here have been addressed in products that are already sampling. You can probably expect to see DMFC fuel-cell battery chargers as well as replacements for notebook batteries in time for next Christmas. But serious market share will have to wait for further refinements in micro fuel cell technology, not to mention price. PD