In the conventional approach to purifying hydrogen, an alloy of 73% palladium and 27% silver is drawn into long thin tubes, about 3 mm in diameter and 6 meters long. Clusters of these tubes are placed inside a vacuum chamber and heated to between 400 and 500° C.

Impure hydrogen gas is then pumped into the small tubes, and the hydrogen readily diffuses through the palladium-silver tube walls and is captured in the outer chamber while the impurities travel out the other end of the tubes.

Hydrogen is tough to handle because of the small size of the atoms and because it naturally wants to bond with other elements. Palladium acts like an atomic filter – the hydrogen atoms readily diffuse right through the metal.

Palladium is $11,000 a kilogram, and even if you didn’t choke at the price, there isn’t enough palladium in the entire world to convert the world’s automobiles to hydrogen power. So the trick is to find a material with the same properties as palladium that is cheaper and much more readily available.

—Alan Russell, Ames Laboratory

A substitute material will need to be less expensive and readily available; allow hydrogen to pass through it; and be ductile enough to be drawn into long, thin tubes. It also has to resist oxidation, because oxygen and water vapor are commonly present in impure hydrogen. And finally the material has to handle repeated heating and cooling cycles, while loaded with hydrogen, without becoming brittle.

The three-year project is being spearheaded by Robert Buxbaum, president of REB Research, a Michigan firm involved in hydrogen filtration and fuel-cell technology (and makers of the “Mr. Hydrogen” purifier). Buxbaum is particularly interested in a membrane reactor which combines hydrogen generation and filtration right at the fuel cell.

Buxbaum obtained $2.8 million from DOE to find substitutes for platinum and palladium. Besides Russell and visiting Chinese scientist Jie Zhang, the project includes Larry Jones, director of Ames Laboratory’s Materials Preparation Center, as well as researchers at Los Alamos National Laboratory, the National Energy Technology Laboratory, and G&S Titanium, an Ohio-based materials fabrication firm.

Buxbaum proposed developing 100 different alloys, relying on the expertise of Russell and Jones in the field of metals development to pick the mixtures.

Using X-ray diffraction technology to study the crystal microstructure of the materials, Zhang can determine whether the materials show promise in terms of ductility. This provides a shortcut of sorts so that the team doesn’t waste time on materials that are potentially brittle. A little more than a year into the project, about 60 binary alloys have been developed with additional ones in the planning stages. The results have been mixed, but Russell indicated one sample is quite promising and several others show promise.

There have been surprises. Some alloys that you would expect to be ductile turn out to be hopelessly brittle, like glass. We also tried a material with 25 percent ruthenium, an element which is notorious for making alloys brittle, but that material turned out to be quite ductile.

—Alan Russell

Samples produced in Ames are first cold rolled to see if they are ductile. Those showing promise are further tested and shipped to REB Research where they’re tested to determine how easily hydrogen will diffuse through the metal. Those showing promise get further testing to see if they can be formed into tubes and how they respond to heating and cooling cycles. But even those materials that are rejected as a palladium substitute, may ultimately wind up as useful for other purposes.