Algorithm works out nanoparticle structure

Since nanoparticles don't have long-range crystalline order, it's hard to use conventional crystallographic methods to look at their structure. But now researchers at Michigan State University, US, have come up with an algorithm that enables them to determine the three-dimensional structure of a nanoparticle from one-dimensional X-ray or neutron diffraction data. "Recreating a 3D nanoparticle of around 100 atoms from a sparse 1D dataset is a computationally challenging problem," Simon Billinge of Michigan State University told nanotechweb.org. "Traditional algorithms such as a standard Monte Carlo approach failed for very small clusters. We decided to try a cluster build up approach: if we could find small sub-clusters that were consistent with the data, we could try to build them up to the final particle." The team used the algorithm to solve the structure of C60 molecules from atomic pair distribution function (PDF) data measured using neutron measurements. Typically, PDF data from a single element system contains a simple list of the atomic distances present in the cluster – it does not provide information on how those atoms are arranged. "[The cluster build-up algorithm] uses a competition between the clusters like a soccer league," said Billinge. "Good clusters get promoted and grow by buying players (atoms); poor clusters are relegated and sell their worst player before competing again at the lower level. The cluster that ultimately wins the top division, the league champion, is the correct cluster. We call it the Liga algorithm [after Spain's La Liga football league]." Although the team also succeeded in getting a genetic algorithm to work, the cluster build up algorithm worked "orders of magnitude better" for determining the structure of nanoparticles. "The most likely applications initially are in inorganic nanomaterials, as the approach is unlikely to be effective for very large and low symmetry molecules like proteins," said Billinge. "It could also be used for studying molecules and clusters in solution. By incorporating chemical information and other a priori constraints it may be possible to study quite complex systems, but we don't yet know how far it will go." Now the team is extending the Liga algorithm to handle multi-element systems and larger clusters, with a view to applying it to "real systems of physical and chemical interest". Billinge believes the algorithm may also be useful for other computationally hard problems. "Source":[ http://nanotechweb.org/articles/news/5/3/19?alert=1].

---

Related articles

Nanoparticles improve delivery of medicines and diagnostics - Tiny, biodegradable particles filled with medicine may also contain answers to some of the biggest human health problems, including cancer and tuberculosis (14.4.2007)