Hi John Burgess (and others),
This is some interesting reflectivity info I snatched while
surfing, :o) n.b: Aluminium is pronounced AL-YOU-MINI-UM
NOT AL-LOO-MIN-UM (Sorry!)
Have a look at http://www.opti-forms.com/optigraph.html for
a neat reflectivity graph of various metals (unfortunately no
Plus from an astronomy type thread:
Because it (silver) reflects more in the visible and IR (but not
UV). Aluminum has a reflectance dip in the near IR. Properly
overcoated silver reflects abouyt 99% from 450 nm to the far IR
with no dips. Silver is much harder to vacuum deposit. It does
not stick well at all to glass or quartz, so some sort of
underlayer (nickel and/or chromium) is normally used. And it, of
course, tarnishes. So an overlayer is needed to protect it.
Fluorides such as magnesium or thorium fluoride are usually used.
These work well, but are more fragile than the quartz coating
used on aluminum.
Silver can be deposited chemically, by several processes, at
least one of which has the interesting side-effect of tending to
produce byproducts that are shock-sensitive exp= losives. Plain
silver tarnishes quickly, so that the coated surface must be
recoated frequently -- in severely polluted areas, perhaps after
only a few weeks of use. (That's only for coatings on the front
of a mirror -- things like the mirror in your bathroom can be
silvered on the back and coated with a layer of nice green paint
that will protect the reflective surface.)
Aluminization, by vacuum deposition, was a great step forward in
telescopes, if only because the aluminum coatings lasted so much
longer. Aluminum also reflects more in the blue, where early
photographic plates were sensitive, so it was particularly
desirable for photographic telescopes. But at longer
wavelengths, it does not reflect as well as silver.
Modern coatings involve layers of various things, overcoated
with layers of protective substances, and can be optimized -- at
cost -- to do better than either silver or aluminum.
(Courtesy of: http://www.polaris.net/~tas/coat.html
:see the address for more of this thread)