November 26, 2003
Before you own a car that's powered by fuel cells, you'll probably have a cell phone or laptop that is. Many engineers now think the first fuel cells to come into wide use will be tiny power sources that replace—and outperform—batteries in wireless electronics. Carnegie Mellon's Institute for Complex Engineered Systems (ICES) is helping to pave the way.
"It's quite an ambitious feat," adds Gary Fedder, Professor of Electrical and Computer Engineering and Robotics. "There is no reason these things can't work. But it will take a lot of engineering to make them work."
The ICES team is working on direct methanol fuel cells, or DMFCs. Most fuel cells run on hydrogen, and thus need bulky storage tanks or extra equipment to produce the gas. DMFCs can be compact because they use a simple liquid fuel: a solution of about 3% methanol in water. Better yet, water is also a byproduct of the cell. If it can be recycled effectively, the cell needs to only store enough H2O to start up. Most storage can be devoted to pure methanol, a little of which goes a long way.
Picture a unit roughly the size of a cigarette lighter: 5 cm. by 3 cm. by 1 cm., with a fuel cartridge taking half that space. This planned ICES prototype would generate 0.65 watt—suitable for a cell phone or basic handheld—for 20 hours. To "recharge," you'd just pop in a refill cartridge. Another prototype, even smaller, would generate 10 milliwatts for more than a month. Potential uses include security: cubes could be placed to detect and pinpoint releases of a toxic substance, for example.
Both DFMCs will use similar technologies to deliver three to five times the output of equivalent lithium-ion batteries. Full-up prototypes are a couple of years away, however.
From microphones to micro-pumps, Fedder is co-director of CIT's MEMS (Microelectromechanical Systems) Laboratory, which builds systems-on-a-chip that include moving parts. One challenge in methanol fuel cells is to circulate and mix water and methanol in the right quantities: you need small, precision pumps and valves. Fedder plans to build them with a novel approach called micro-membrane technology.
"We start with a CMOS chip," Fedder explains. "In the cleanroom at Hamerschlag Hall, we etch the chip"—first etching the holes down through the upper layers to produce a grid pattern, then etching away at the pure silicon substrate beneath. This leaves a thin metallic mesh suspended over a dished-in cavity, and held in place along the edges. Finally, the holes in the mesh are sealed with a polymer. You now have a membrane that can be made to flex up and down like a drumhead."
In fact, the micro-membranes were invented for acoustics by Professor Ken Gabriel's group at Carnegie Mellon.
"Our concept," says Fedder, "is to adapt it for pumps and valves that push liquid through our system by peristaltic action." Through a channel, liquid will flow into the cavity under the membrane. When the membrane is flexed down by a control signal, it will squeeze the liquid out through another channel.
A DFMC begins to work when methanol (CH3OH) and water come in contact with a catalytic surface at the anode, one of the most important components of the fuel cell. A chemical reaction is triggered, breaking the fuel into CO2, hydrogen ions (protons) and free electrons. The protons migrate to the cathode through an internal membrane that does not let the electrons pass. And that produces a difference in charge across the cell, pulling the electrons through an external circuit to generate a current.
Fedder estimates that a DMFC unit for a handheld will cost $100, though the refill cartridges should be only about $1 each.
Researchers worldwide are working on DMFCs. Some firms are showing rough prototypes: Toshiba has one for a laptop. But the game is still early stage and Carnegie Mellon is poised to play a major role.
"Our main strength lies in integration," believes Christina Amon, the Director of ICES and the Raymond J. Lane Distinguished Professor of Mechanical Engineering. By that she means both physical integration—such as with MEMS techniques for making micro-scale pumps, separators and other fluidic elements all from one piece of silicon—and also multidisciplinary integration.
"ICES exists to help pull together the strengths we have in materials, computer modeling, and all the disciplines," Amon says. "That's essential in fuel cells. We are planning to grow in fuel cells."
While ICES research focuses on micro-fuel cells, some CIT students are helping to develop the largest of the breed: the solid oxide fuel cells meant for power plants. Juniors, seniors and grad students in the 600-level ICES project work with Siemens-Westinghouse, a leader in solid oxide technology, on a range of modeling and design issues. David Archer, Adjunct Faculty in Mechanical Engineering, is coordinator.
Excerpted with permission, Carnegie Mellon Engineering News, Fall 2003. Text by Mike Vargos