Re: ATM Determining the index of refraction of glass

[Anthony Stillman <atmer@flash.net>]

>...is it possible to determine the refractive index



Chuck, others,

Your request is most timely.  This last weekend I finished the upgrade to
my direct measuring jig.  Actually I'm working on higher percision
measurment aperatus both for direct measure and otherwise.  This jig uses
an old optical depth micrometer and colored lights.  The next jig uses a
Unitron inverted microscope and a digital micrometer, but its still needs
some work.  The high percision device uses a glass V-block to hold the
sample and a really long lever arm to measure the deflection.  Though that
design can measure index accurately, it will measure dispersion far more
accurately.

In the current incantation I focus on and through the glass on marks on
either side.  One side should be polished.  I tried using an index matching
fluid to measure a plano-plano but ground piece of glass and found the
readings unreliable.  I'm using Radio Shack LEDs as light sources.  I
calibrated the peak wavelengths in a monocrometer and then I calibrated
them again after calibrating the monocrometer.  Peter John Smith suggest I
use a cold cathode florescent as he does.  I think that's a good idea and
try it soon.  The LED peak emmisions were all actually quite close to what
was printed on the packaging.  There is some variation from LED to LED.

Direct measurement of index is only good to about 3 places, maybe a little
bit more.  With the specific gravity this is good enough to ID a common
glass, but it can't tell the difference between BK7 and UBK7 and it
certainly isn't enough for a triplet.  Starting with only three places,
even a doublet will likely need a little tweeking.

Basically the process is pretty simple.  I first glue the glass in place
under the optical micrometer, polished side up.  Well, OK, first I position
the glass so the marks I will focus on are in the field of view, then I
glue the glass in place.  Next I focus on the bottom mark while its
illuminated with red light (697nm) and record the reading.  I repeat this
several time.  Then I change the light to orange (620nm) and do the same
thing again.  Ditto with yellow (595nm), yellow-green (565nm), and blue
(430nm).  Next I focus on the top mark and repeat the whole process over.
The differences in the readings are the optical depth of the glass at the
various wavelengths.  I have to take the differences as the optical depth
micrometer is refractive and hence also effected by the wavelength.  The
physical depth devided by the optical depths are the refractive indices at
those wavelengths.  To measure the physical depth I am using a Brown &
Sharpe micrometer good to 0.00005"  To finish I plug the indices into a
non-linear least square fit spreadsheet I built to give me the Cauchy
parameters and then plug these parameters into yet another spreadsheet to
give me the indices at the standard wavelengths.

Some things I've learned along the way.  First, my eye isn't so good at
focusing red light.  On the other hand the CCD camera I taped in place was
worse.  Second,if you want three places the sample can not move, hence the
glue.  Third, ultra-bright blue LEDs are actually florescent and have a
broad emmision profile, though far and away most of the light is at 430.
Fourth, it helps bunches to have a know glass standard.  Sixth, more is
better.  My optical micrometer is good to 0.00005 inches if I use it
carefully and repeat, repeat, repeat, use the highest power I can and am
more rested than tired.

I'm toying with the idea of hauling the jig along to Riverside, to measure
old ball glass that could be used if only it was known, but only if there
is interest.  Its very time consuming.  One side must be polished at least
where the mic goes and optical part of this jig tops out around 1.5" of
physical depth, depending on index.  My B&S mic tops out at 1"

Anthony