NASA NTRS2004
The original motivation for the construction of the BESL prototype was to obtain the first proof of a 50-year-old hypothesis regarding the solidification of liquid metals. Since the 1950s it has been known that under proper conditions liquid metals can be cooled below their melting temperature (undercooled) without crystallizing to the stable solid phase. In 1952 Frank proposed that this was because the atoms in the metallic liquid were arranged with the symmetry of an icosahedron, a Platonic solid consisting of 20 tetrahedra (4-sided pyramid-shaped polyhedra) arranged around a common center. Since this local atomic order is incompatible with the long-range translational periodicity of crystal phases, a barrier is formed to the formation of small regions of the crystal phase, the nucleation barrier. A proof of Frank's hypothesis required a direct correlation between measured icosahedral order in the undercooled liquid and the nucleation barrier. The tendency of sample containers to catalyze nucleation obscured this relation, requiring containerless techniques. Combining containerless processing techniques for electrostatically levitated droplets (ESL) with x-ray synchrotron methods, a team from Washington University, St. Louis, MO, NASA Marshall Space Flight Center, and MUCAT at the APS demonstrated an increasing icosahedral order in TiZrNi liquids with decreasing temperature below the melting temperature. The increased icosahedral order caused the transformation of the liquid to a metastable icosahedral quasicrystal phase, instead of the stable tetrahedrally-coordinated crystal intermetallic, giving the first clear demonstration of the connection between the nucleation barrier and the local structure of the liquid, verifying Frank's hypothesis for this alloy.