A radioactive topic

released 08.25.09

By J. William Bell

Materials used for radioactive-waste storage must be especially resilient. NCSA and TeraGrid resources help researchers identify and understand next-generation candidates.

Some 60,000 tons of high-level radioactive waste is being stored in the United States, according to a 2006 report from U.S. Geological Survey scientists. Coming largely from nuclear power plants and retired nuclear weapons, it resides at more than 100 temporary sites around the country.

The containers used today aren't viable candidates for long‑term storage. They're 100-year solutions to 100,000- or million-year problems, based on "a special glass that cracks. Not very good for long time periods," explains Jianwei Wang, a post-doc at the University of Michigan.

Finding the right material in which to store the waste underground over the course of millennia is a great challenge.

As part of a team from Michigan and Rensselaer Polytechnic Institute, Wang is using TeraGrid resources to investigate pyrochlores. These minerals show natural resistance to radiation and can be altered to make them hundreds of times more resistant to the structural defects that crop up under the extreme irradiation and pressure.

With NCSA's Cobalt system, the team is showing that the performance of these materials in extreme environments is directly related to the energies at which defects form. They also use the Kraken system at Oak Ridge National Laboratory.

For example, their studies show that two pyrochlores—gadolinium titanate and gadolinium zirconate, in which titanium can be exchanged with zirconium—display dramatically different behaviors under irradiation and pressure. Results were published in Physical Review Letters in 2008. Modeling the pyrochlores at progressively higher pressures and 1,800 Kelvin showed that the zirconium-based structure could accommodate 1,000 times more defects than the titanium-based structure at pressure up to 20 gigapascals.

The modeling also showed that the number of defects in the zirconium-based structure decreased while those in the titanium-based structure increased as the pressure increased toward 20 gigapascals. These results were presented at the Materials Research Society Fall Meeting in December 2008.

Perhaps counterintuitively, defects at the atomistic scale improve materials' resistance to radiation, according to Wang. "Irradiation induces defects and the accumulation of these defects can lead to disorder in the material. But studies show that irradiation resistance improves in proportion to the atomic disordering tendencies" of the materials.

The team combines their computational simulations with ion beam experiments. Rod Ewing, a Michigan professor who leads the group along with Michigan's Udo Becker, says the two approaches "work well together. They're intimately mixed."

Experimental findings validate computational findings, and vice versa. Just as importantly, they show the researchers different parts of the whole. "Experiments can look at the gross effect [that radiation and pressure have], but to gain insight into the process itself, we need simulation," Ewing explains.

At the fundamental level, computations show how the free energies of atoms of materials with different chemical structures vary under more and more extreme conditions. Those differences simply can't be observed in the ion-beam experiments.

Overall, this combination of experiment and computing "provides a scientific basis for the next generation of nuclear waste [containers]," says Wang. "The computational part of the effort is important for the success of our projects, and the TeraGrid resources are essential for our computational work."

"Not to be immodest, but this is exciting and important science. It allows us to begin to consider ways we can make materials that are more radiation resistant," Ewing says.

Team members
Rod Ewing
Udo Becker
Jianwei Wang
Fuxiang Zhang

Funding
Department of Energy
National Science Foundation