Senior Author

Complex inorganic nanoparticles are of great interest to materials scientists. But learning about their structure is difficult, because X-ray crystallography often fails on these tiny molecules. Researchers at Michigan State University in East Lansing decided to see if they could combine one-dimensional data with mathematical modelling and computation to predict the three-dimensional structure.

They reasoned that if they measured the distance between pairs of atoms, they could use computers to work out the matching three-dimensional structure. On page 655, Simon Billinge and his colleagues present an algorithm based on UK soccer leagues that seems to be able to work out the structures.

How does the soccer analogy apply to atomic structure?

In soccer, you have promotion and relegation. Teams that win get promoted to a higher level and those that lose get relegated to a lower level. Relegated teams earn less money and sell their worst player, whereas promoted teams can buy new players. The atoms are the players, and clusters of atoms are the teams.

Where did you get this inspiration from?

The basic ideas grew out of conversations about soccer with my postdoc, Pavol Juhás. I guess we spent too much time watching European soccer growing up.

What's this method's biggest advantage?

Relegation gives you a way to backtrack — it turns out that this algorithm is really efficient. Maybe soccer teams could use it.

What's new about this?

We demonstrated for the first time that it is possible to get unique three-dimensional nanoparticle structures from one-dimensional diffraction data.

What's the next step?

This was a proof of principle. All that we've solved is the buckyball — a structure that's already known. We need to use this approach to solve a more complex, unknown nanoparticle structure.

What football team do you support?

I have supported Arsenal since I was a boy and right now they are playing the most beautiful football.

What molecule do you support?

It would have to be buckyball. Because in this case, buckyball won the league.

What other British pastime could you use to create algorithms?

Maybe our next algorithm could be inspired by discussions of the weather.