In some ways the findings fly in the face of previous assumptions. Platinum, known for being one of the best catalysts for promoting the combination of hydrogen and oxygen in fuel cells, was one of the first materials tried for catalyzing lithium and oxygen in lithium-air batteries. But experiments showed that it actually did a poor job, so platinum was dropped.

The MIT researchers found that platinum is useful in lithium-air batteries, but for the opposite reaction–freeing oxygen from lithium oxide during charging. “Everyone knew that platinum was inactive for discharging the battery, but we showed that platinum was one of the best catalysts for charging,” Shao-Horn says.

On the other hand, gold is typically considered a poor catalyst because it is inert, Shao-Horn says. Indeed, the MIT researchers had first used gold as a sort of control in experiments to measure reactions involving a poor catalyst. To their surprise, they found that gold does a good job of catalyzing the combination of lithium and oxygen–much better than platinum. (Toyota researchers had shown this previously, and issued a patent a few months before Shao-Horn’s group saw the effect.) Furthermore, the researchers found that both catalysts became more effective when they were combined as nanoparticles. “Together they work synergistically,” Shao-Horn says.

In addition to improving efficiency, promoting these reactions could also potentially increase the number of times that lithium-air batteries can be recharged, by minimizing the accumulation of lithium oxide, which otherwise clogs up the battery. As they continue to develop lithium-air batteries, the MIT researchers will explore this possibility; they will study the gold platinum catalysts in more detail to understand how they work; and develop new catalysts with different combinations of materials.

The MIT researchers are also working to cut the cost of the catalyst by using less platinum and gold. One option is to coat nanoparticles made of cheaper materials with thin layers of these precious metals. Other researchers have demonstrated that inexpensive manganese oxide catalysts can be effective for lithium-air batteries, says Jean-Marie Tarascon, a professor at the Universite de Picardie Jules Verne in France. He says that this material recently has been shown to produce even higher efficiencies than Shao-Horn’s catalysts, although that data is not yet publicly available.