Re: How does supercooling exact;y work? Does it work on all liquids?

First let’s consider pure liquids.  (Solutions cooled below the point at
which precipitation occurs are sometimes called supercooled but are perhaps
more properly called supersaturated.)  In principle any liquid can be
supercooled.  In general the more viscous a liquid and the weaker the
intermolecular forces that make it a liquid the more easily it can be
supercooled.  Because of its low viscosity and strong hydrogen bonding
water is actually rather difficult to supercool.  Supercooling produces an
unstable state where one phase of a material is at a temperature below that
at which it would be in equilibrium with some other phase of the same
material.  Typically when a liquid is cooled to its normal freezing point
some of it will change to the solid form and the temperature will stop
decreasing as energy is removed and instead the rest of the liquid changes
to a solid.  But change takes time, so sometimes…  if you don’t cool it too
slowly, or too rapidly, and there aren’t any suspended solids (dust etc.)
or too many scratches or dirt on the container and you don’t shake it too
much you can produce a container of liquid that is at a temperature below
its freezing point.  It may then stay in this state either very briefly or
for quite some time before eventually solidifying rather abruptly.  This
solidification can also be triggered by the addition of a small amount of
the solid form of the liquid (a seed crystal), shaking, which produces
either bubbles or cavitation that act a nucleation sites, or even a speck
of dust or scratching the wall of the container.  You will sometimes hear
that glass is a supercooled viscous liquid.  For a good presentation of why
this is probably not the best description for glass see Zanotto in the
American Journal of Physics 66(5)392-395 1998.  You will also sometimes see
articles about supercooling a solution (one thing dissolved in another)
below the point at which precipitation would normally occur. (The
supersaturated solutions mentioned above)  These exhibit behavior similar
to a supercooled liquid except the liquid stays liquid even after all the
solid has formed unless it becomes trapped in the crystals formed.  For a
good description of how to do this yourself see the page 80 of the October
2007 issue of Popular Science.  You might still be able to bring this up
from www.popsci.com under their How 2.0 or DIY sections.