Five-Fold Symmetry in Liquids
Entry by Emily Redston, AP 226, Spring 2012
Five-fold symmetry in liquids by F. Spaepen. Nature 408, 781-82 (2000)
Even though liquids are essential ingredients in soft matter, researchers have only recently begun to understand their structure. Previous theories describing liquids as disordered crystals or as dense gases have fallen short, and it has generally been accepted that the liquid structure is a well-defined phase in its own right. This was demonstrated in the late 1940s when Turnbull showed that many simple liquids could be cooled far below their freezing points without crystallization occurring. This is only possible if the liquid structure is fundamentally different from that of a crystal. While there are still many unanswered questions, the current method is to explain the liquid structure as a dense packing of tetrahedral building blocks. Ideally one would like to have a simple structural method for describing the liquid structure as we already have with the periodicity of crystal structures. Unfortunately, since the atomic structure of a liquid changes over time and space, conventional scattering experiments using X-rays or electrons or neutrons can only provide directionally averaged information.
Reichert et al  were the first to report direct evidence for polytetrahedral structures in a monatomic liquid trapped at a solid interface; they saw the characteristic five-fold symmetry of the bonds. Typical crystalline solids (like the face-centered cubic structure) are formed by maximizing long-range order. The polytetrahedral packing in liquids, on the other hand, maximizes the short-range density of the structure. The densest local configuration that can be created with hard spheres is a tetrahedron. Five tetrahedra can be packed around a common edge, but they leave a gap of about 7° (Fig. 1c). The thermal disorder of liquids, the ease by which they flow, and the rapidity of atom diffusion can be qualitatively understood by the redistribution of these 7° gaps. Tetrahedral packing leads to five-fold symmetry, which is incompatible with long-range periodicity. Thus polytetradral short-range order favors disordered or amorphous structures. The best model for explaining the scattering data from liquids is Bernal's dense random packing of hard spheres.
A complete understanding of the structure of simple liquids is extremely important and relevant to a lot of work done in soft matter. Currently we know that liquids contain many configurations with five-fold symmetry. I find it a little surprising that something that seems so simple turned out to be such a complicated problem. I'm very interested to see where our understanding will be down the road, and whether we'll ever have the complete picture with liquids that we have with crystalline solids.
 H. Reichert, O. Klein, H. Dosch, M. Denk, V. Honkimäki, T. Lippmann, and G. Reiter, Nature 408 839-841 (2000)