Platinum and most other noble metals are very good catalysts. Although gold was once thought to be the exception, scientists discovered in the 1980s that gold particles smaller than 5 nm were more active than other noble-metal catalysts for reactions like low-temperature carbon monoxide oxidation. In contrast, platinum nanoparticles smaller than 5 nm are less catalytically active. Until now, no one knew why this was the case, but a team of researchers at Oak Ridge National Laboratory have come up with an answer.

Calculations by team member Sergey Rashkeev (now at Idaho National Lab) show that the key difference between the behaviour of gold and platinum at the nanoscale is that although both metals bind reactants more strongly as the particle size becomes smaller, in platinum the binding becomes so strong that the reaction never proceeds at low temperatures. In gold, however, the weaker binding and flexibility of the nanoparticles allows catalytic activity to continue.

"Bulk gold is not very reactive, which is why it is so good for jewellery," team member Andrew Lupini told "but, low coordination sites on gold nanoparticles become much more reactive. Moreover, the gold nanoparticles become softer and more flexible at these small sizes, allowing the reactants to move and interact." He adds that it is this combination of being able to bind to reactants, while still allowing them to interact with each other that allows gold to be active on the nanoscale.

Lupini and colleagues employed a special aberration-corrected scanning transmission electron microscope (STEM), which held the record for the highest resolution image until very recently. The researchers were able to obtain images of gold clusters that showed them to be smaller and thinner than previously observed (see figure). They then used an extensive series of first principles calculations to determine the activity of sites with different coordination and how the number of such sites varied with particle size.

Applications for gold nanoparticles catalysts include respirators, reducing pollution in automotive applications and selectively removing carbon monoxide from hydrogen gas in fuel cells.

The team will now investigate novel ways of making such catalysts. "By combining catalyst synthesis and reactivity measurements with high-resolution STEM and first principles theory in one synergistic program, we hope to gain new fundamental insights into how they work at the atomic scale so that researchers can produce new catalysts for more efficient reactions and maybe cleaner, greener fuels in the future," says Lupini.

The researchers reported their work in Physical Review B. The research was sponsored by the US Department of Energy's Office of Basic Energy Sciences.