Mechanisms of melting!

Melting and solidification are probably the most ubiquitous temperature driven phase transformations. As we increase the temperature of a solid, at a particular temperature known as the melting temperature, the solid becomes a liquid. On the other hand, if we keep decreasing the temperature of a liquid, at a particular temperature known as the freezing temperature, the liquid becomes a solid. However, under some circumstances, it is possible to keep a solid above its melting temperature in the solid state: this is known as superheating. Similarly, keeping a liquid below its freezing temperature in the liquid state is known as supercooling.

The assumption that melting starts at the interior amounts to assuming homogeneous nucleation of melting; surfaces, or grain boundaries, if present will be spots of easy melting, and in such cases the melting is said to nucleate heterogeneously. In fact, the fact that the surface melts easily is what leads to a size dependence of the melting temperature, which is being widely studied by the nanomaterials community these days.

Thanks to Deep for pointing out that (a) what Born and Lindemann proposed are criteria; and, (b) unless extreme care is taked to avoid it, melting nucleates, almost always, heterogeneously.

The mechanism of melting, however, is yet to be clearly understood. Consider a crystalline solid and assume that the melting starts in its interior. As temperature of the crystal increases, its shear modulus (and hence its ability to withstand non-hydrostatic stresses) decreases. Melting could then be that point where the shear modulus becomes zero. This shear-induced melting criterion was proposed by Born. As the temperature increases, the atoms in their lattice positions start vibrating more and more violently. Finally, at the melting temperature the vibrations are so violent that the crystalline lattice structure breaks down leading to melting; this criterion was proposed by Lindemann. The third criterion involves the production of defects in a solid with increasing temperatures. At the melting point, the defect concentration is so large that it leads to the destruction of the crystalline lattice, and hence to melting. Over the years, there had been several studies which discuss these different mechanisms. Deep referred me to this recent article in the May, 2005 issue of Nature Materials on some molecular dynamics studies of melting in superheated crystals. Here are some pointers to some of the earlier literature on these different mechanisms of melting. Finally, here is a nice post by Santonu on defects in solids, albeit with an emphasis on segregation.