Fuel-Cell Tanks Buck Convention

Mark Anderson 05/16/05

On the road to petroleum independence and greenhouse-gas reduction, the old internal combustion engine will someday need to be scrapped. That will only occur when the alternative -- most likely the hydrogen-powered fuel cell -- is as cheap and convenient to use as the conventional automobile is today.

It's perhaps the toughest challenge in fuel-cell research: designing a safe, lightweight and compact hydrogen fuel tank. Two recent papers, published in the April 22 and May 6 issues of the journal Physical Review Letters, find the most promising hydrogen storage medium is in something called carbon nanostructures.

Just as the semiconductor revolution became possible when wafers of silicon were doped with other elements, the hydrogen revolution could be realized with the help of tiny balls or tubes of carbon decorated with periodic defects in their structure. Those defects are metallic elements that attract hydrogen to the nanostructures like Velcro. In the May 6 paper, the hypothesized fuel tank would be filled with carbon nanotubes coated with the metal titanium. The April 22 paper proposes a structure involving carbon buckyballs and another metal such as scandium.

In both cases, the experience of gassing up the car would be similar to today, although the microscale goings-on would be markedly different. The driver would pump hydrogen gas into a tank of comparable size to automobile gas tanks today. The grid of coated nanotubes or buckyballs inside the tank would then soak up the hydrogen, incorporating the fuel into the structure of the tank itself. When the car was turned on, the engine would suck on the hose connecting to the tank, causing the hydrogen to dislodge from its temporary storage and float into the fuel cell to be converted into electricity and water vapor. (The nanotube tank would also require some extra heating to coax the hydrogen into the engine.)

The upside is that the tank would theoretically hold enough hydrogen to power the car for as much as 300 miles of driving. And the gaseous hydrogen would never be at high enough pressures or concentrations to cause any safety concerns. No Hindenburg issues here.

The downside is that both ideas are only in the most preliminary of phases. Depending on how the actual chemistry, physics and engineering play out, that 300 miles between refueling stops could in reality turn out to be only 150 or 100 miles. And the price of such a car is anybody's guess.

"I'm not absolutely sure that hydrogen will ever be exactly the same as gasoline," said Anne Dillon of the National Renewable Energy Laboratory, one of the co-authors of the buckyball paper. "People might have to relax a little bit and refuel a little more often to save the planet."
Both papers, funded by grants from the Department of Energy, are theoretical and computational simulations. Neither team has made the nanostructures envisioned, although both have now taken their quest to the lab to try fabricating the materials proposed.

"This system is so clean -- it's carbon, hydrogen and a very simple metal," said Taner Yildirim of the National Institute of Standards and Technology, co-author of the nanotubes paper. "So the theory is very powerful."

Dillon, an experimental chemist in a team of theoretical physicists, relishes the challenge ahead. "This is the first time in my career where I've finally thought that this could work," she said.

"It makes physical sense. The physicists believe it, and the chemists believe it. Physicists can predict something, and a chemist can look at it and say, 'No way in hell can I make this.' But this time, I feel like it might be possible."

Regardless of how feasible these latest two proposed nanostructures ultimately prove to be, Mildred Dresselhaus of MIT noted that both teams are pioneering a promising new field that could ultimately enable a hydrogen-fueled future.

"What we have said in the past is that some kind of defect in a nanostructure would be the right way to go," she said, "because both (the defect and the carbon nanostructures) are needed for changing the equation."

Disponible en: http://www.wired.com/science/planetearth/news/2005/05/67512