Computer-Predicted Catalysis

The basic theory of how chemical reactions happen–molecules approach each other, overcome potential energy, and then form new reactants–has held up in experiments almost every time. But the theory doesn’t fully explain what happens when a molecule approaches a metal surface, such as the surface of an industrial catalyst. This is important because metal catalysts are widely used in catalytic converters, fuel cells, and even to make margarine.

What makes metals tricky is that they don’t have discrete energy states like molecules–rather than jumping from one specific energy level to another, electrons move between energy states in a metal in a more continuous way.

Two papers published in the journal Science this week use new algorithms to better describe what happens at the surface of metals including catalysts.

One describes the interactions between a gold surface and nitric oxide molecules excited using a laser. Older models predict that when the gas hits the gold surface it will still be vibrating. The new model predicts what actually happens: the molecule electronically couples to the gold.

The second paper looks at the interaction that cause hydrogen atoms on a copper surface to bond with one another and form hydrogen gas. It remains to be seen whether these results can be generalized, but if they can it could lead to a better understanding of the metal catalysts widely used in industrial chemistry.