Amorphous structures from diffraction data?

Despite the importance of disordered materials, we still find it very difficult to determine their structures, because the lack of long-range prevents the use of conventional crystallographic techniques. This ’nanostructure problem’ as it has been termed, is one of the biggest challenges in the solid state sciences. In this paper we show that by making use of the idea of ‘invariant’ local environments in structure determination - in particular by minimising the variance in local data - it is possible to make significantly better models of amorphous and nanostructured materials.

Figure: three sections from configurations of amorphous silicon. From left to right, the RMC model, our INVERT model, and a good model of amorphous silicon (made using the WWW algorithm). The INVERT constraint produces a much more realistic configuration, as can be seen by the lower number miscoordinated atoms (red) compared to correctly coordinated (blue).

A viewpoint article featuring this work is available at Physics, which provides very important context on both the importance of this problem and possible solutions.