melting-point depression

Everyone learns that a material has a fixed melting point — gold melts at 1064 degrees Celsius, full stop. That is true for a lump you can hold. But shrink gold into nanoparticles a few atoms across and it starts to melt at far lower temperatures, sometimes hundreds of degrees below the textbook value. The melting point is not a fixed constant of the material after all; for very small crystals it slides downward as the crystal gets smaller. This sliding is called melting-point depression.

The reason comes straight from the surface. Atoms on the surface of a crystal have fewer neighbours holding them in place than atoms buried inside, so they are more loosely bound and easier to shake loose — melting always begins at surfaces. In a big crystal the surface is a tiny fraction of the whole, so it barely moves the melting point. But in a nanocrystal a large fraction of atoms sit on or near the surface, so the whole particle is easier to melt. Thermodynamics makes this quantitative: the melting temperature of a particle of radius r follows roughly T_m(r) = T_m(bulk) times (1 minus C over r), where C is a small constant set by the surface energies — a relation of the Gibbs-Thomson type. The 1 over r means the depression is negligible for large r and grows sharply as r shrinks into the low nanometres.

This is not a laboratory curiosity — it governs how nanomaterials are made and used. Nanoparticle inks and pastes can be sintered (fused together) at gentle temperatures a plastic substrate can survive, precisely because the tiny particles melt or coalesce early. It sets limits on how small and how hot nanoscale devices can be pushed before they degrade, and it is a clean, honest illustration of the field's central theme: at the nanoscale, surface energy is not a small correction — it rewrites even something as basic as when a solid becomes a liquid.