For each of these solutions, various salts, sugars, alcohols, etc in
the starting solution don't fit well in ice, and are almost completely
excluded from the ice. The basic pattern- lowering the freezing point
by adding the solutes- is the same for each.
My great-grandfather used this principle for an important practical
application during Prohibition. He could brew some hard cider, but it
didn't have a very high alcohol concentration. Leaving it out in the
Chicago winter made a lot of the water freeze out as ice, which was
easy to throw out. The liquid left behind had a high concentration of
alcohol and sugar.
The behavior of different components may be different. For example,
the saltwater and sugar-water are solutions, where the salt and sugar
mix in as individual ions and molecules, respectively. For the milk,
however, it is a "suspension", with some components (like salt and
sugar) dissolved in, and others (like globs of milk fat) floating
around, perhaps with some layer of emulsifiers. When frozen, the
different parts may separate out with different fractions.
This separation depends also on how the ice forms. We've answered
a few questions about air bubbles in ice -- you can get the same kinds
of effects by freezing a container of one of these fluids "outside in"
(ice forming on the surface first) -- this may trap the impurities in
the middle of your frozen lump because they cannot escape.
In general, though, the freezing/melting point will be lower, and
will depend on how many molecules of dissolved stuff there are compared
to how many molecules of water there are. Salt counts as just under two
for each NaCl, because the Na+ and the Cl- ions separate (except at
high concentrations, when some of them are still stuck together). Big
globs of fat won't change the melting/freezing point because they don't
interfere with the water molecules forming ice, as they aren't
Mike W. and Tom J.
(republished on 07/25/06)