By Laura Buenning

Touted for their promise as a power source for tomorrow’s automobiles, hydrogen fuel cells still have a way to go to live up to their “clean green” potential.

At the heart of the question lies uncertainty about how the molecular structure of the fuel cell and its energy-producing processes interact to help or hinder the cell’s efficiency, says Myvizhi Esai Selvan, a chemical engineering graduate student with the UT Computational Materials Research Group (CMRG) and recipient of the 2007–08 Institute for a Secure and Sustainable Environment stipend.

If only we could journey to the center of a fuel cell to see exactly what happens there, rather like the miniaturized surgical crew who enter the brain of a wounded Russian defector in their microscopic submarine in the film Fantastic Voyage.

Such a voyage is still a fantasy, but CMRG-created computerized simulations of molecules in action may offer the next best thing. The molecular-dynamics model permits a nanoscale view of the structures that form at the borders where protons enter and leave a polyelectrolyte membrane.

We know generally how fuel cells work. Pure hydrogen enters the anode side of the cell and, aided by a platinum catalyst, splits into electrons and protons. On the other side, a platinum-coated cathode attracts oxygen from air to its surface, creating a kind of electrical pressure called a “potential.”

Separating the anode and cathode, a thin semipermeable membrane—termed a polyelectrolyte or proton exchange membrane (PEM)—allows the protons to pass but rebuffs the electrons and the hydrogen and oxygen atoms. The electrons, attracted to the oxygen on the other side of the membrane, flow into an external circuit around the membrane and back inside the cell, supplying power for electrical motors or other equipment along the way. To complete the circuit, the protons pass through the membrane and pair off to bond with the oxygen atoms and electrons on the cathode side of the cell to form water.

“People have analyzed what’s going on with the electrodes [anode and cathode] and the electrolyte [the membrane], but we don’t have a clear picture of what happens in the interface [or common boundary] between the platinum catalyst on the electrode and the membrane,” Esai Selvan says.

“We want to know how the protons leave the catalyst and enter the proton-exchange membrane,” she says, “and how water influences the process.”

The question has stumped scientists for more than 40 years, says CMRG group leader David Keffer. Keffer and Brian Edwards, UT chemical and biomolecular engineering associate professors, direct the computational materials group.

“We’re trying to determine whether wetting helps or hinders or makes little difference to the electrodes, catalyst, and proton movement through the cell,” Esai Selvan says. This information will improve the durability and cost of the membrane as well as the placement of the electrode-catalyst membrane within the cell.

Keffer, Edwards, and the CMRG can combine what they learn about the chemical engineering aspects of fuel cells with the findings of their colleagues in the Chemistry Department. Chemistry’s Distinguished Professor Jimmy Mays and Professor Mark Dadmun are working to improve polymers suitable for polyelectrolye exchange membranes.

Safe Transport

Fuel-cell design represents one side of the equation; hydrogen storage and transport, the other.

Lightweight, chemically active, and highly flammable, hydrogen is as hard to contain as the best escape artists. Only high pressures can condense it into useful quantities, and cryogenic temperatures are required for it to go into a liquid state. So unless we come up with a way to safely store hydrogen in reasonably lightweight containers, its use as an alternative fuel in automobiles will be limited.

“Storage materials that physically—as opposed to chemically—capture hydrogen have potential,” says Sandeep Agnihotri, assistant professor in UT’s Department of Civil and Environmental Engineering. Under the right conditions, pores inside and outside a material become containers for hydrogen gas or other molecules.

“We use a material’s gas adsorption [the capacity of a material to attract and hold molecules on its surface] to calculate its porosity. The more pores, or holes, a material has, the more surface area where the molecules can cling,” Agnihotri says. And adsorption of hydrogen is reversible; simply remove the conditions holding the hydrogen in place and it becomes available as fuel.

Enter the Carbon Nanotube

Made exclusively from carbon atoms, nanotubes are 1-atom-thick sheets of graphite rolled into seamless cylinders 1/10,000 to 1/100,000 of the thickness of an average human hair yet 100 times stronger than steel. Nanotubes have garnered attention as a possible storage medium for hydrogen because they rarely react with other elements and have both inside and outside surfaces.

“But,” Agnihotri says, “if you track studies on hydrogen storage in nanotubes over the last 10 years you find inconsistent results, and the question is why?”

To find out, he needed a model that told him theoretically how much hydrogen the tubes should be able to carry—a task complicated by the fact that tubes come in many sizes.

“A bundle of nanotubes will typically have diameters anywhere from 7 angstroms (0.7 nanometer) to 20 angstroms,” he says, pulling pens, markers, and pencils from his desk drawer and holding them between his thumb and forefinger to demonstrate.

With the help of Raman spectroscopy to tell him the actual diameters and percentages of each in a particular bundle of nanotubes, Agnihotri calculated the inside and outside adsorption potential for each diameter. By adding them together at the percentages found in the sample, he was able to arrive at a total adsorption figure for the entire bundle.

“But my calculated adsorption was much higher than experimental data from similar samples,” he says.

As he thought about it, Agnihotri began to question his assumption that 100 percent of the carbon tubes would be open to allow the gas inside. Playing with the idea, he found in one case that his adsorption calculations matched experimental data if only 45 percent of the tubes were open and 60 percent in another. In one sample nothing was open.

“Knowing what fraction of the tubes are open will help clarify experimental data,” he says. “Think about the implications for making and processing nanotubes, let alone for accurately predicting hydrogen storage capacity.”

Tags: Biofuel • Brian Edwards • David Keffer • Energy • Engineering • Environment • Modeling • Myvizhi Esai Selvan • Nanotechnology • ORNL • Physics • Sandeep Agnihotri • Technology