Index Home About Blog
Message-ID: <39AD45B9.1A256B2E@bellsouth.net>
From: John De Armond
Newsgroups: rec.crafts.glass
Subject: Dichroic glass
Date: Wed, 30 Aug 2000 13:34:49 -0400

"Peter A. Neary" wrote:
>
> Sorry if this is a bit hair-brained.  But my father seems to think he should
> be able to make dichoic glass using a tungsten filament to vapourize the
> coating material in a vaccuum chamber.  Is he right? If so any information
> would be greatly appreciated.  What coating material/ oxides?  Vacuum
> pressure?   Does the vacuum chaber need to be heated? Why?
> Thanks, Peter.

Tungsten is a poor choice.  Hard to control.  A good metal for your
first experiment would be titanium.  Wire is available from Strem
chemicals (http://www.strem.com).  Titanium yields very pretty
rainbow coloration, even in one coating.  Of course, it will take some
refinement and experimentation to achieve the nice solid colors
that commercial thin film outfits produce but it's doable.

I've done thin film deposition in my shop using little more than a
modified pressure cooker, a low voltage power supply and a high
vacuum system. The pressure cooker is modified by sealing all
openings using high vacuum techniques (welding preferred),
substituting a Viton O-ring for the gasket (lower outgassing at
vacuum), adding a view port (cut a hole, high vac epoxy a  hunk of
thick glass over it.) and bringing two power leads in through
homemade vacuum feedthu fittings (compression fittings with wire
potted in the opening using high vac epoxy.)

The high vacuum stuff (mechanical pump, diff pump, ion or cold
cathode gauge, fittings, etc) can be had used at Duniway Stockroom,
http://www.duniway.com.  Figure $2k for this stuff.

Here's roughly the procedure.

Attach some titanium wire to the electrode on the pressure cooker
lid.

Chemically clean the glass using lab detergent (alconex
recommended), then caustic soda followed by conc nitric acid.  Bake
@ 250 deg F for a couple of hours.

Place in pressure cooker, held off the bottom a half inch or so with
glass beads or some other heatproof, non-porous substance.

Close up the chamber and start the vacuum system.  Optionally, heat
the chamber on a hot plate to oh, 200 deg or so.  Not too high or
the O-ring outgases.  You should achieve 10 E-5 Torr vacuum.  If
not, look for leaks, outgassing components, etc.

Valve off the vacuum chamber and bring the pressure up to about 3
Torr using dry argon.  Establish a plasma discharge in the chamber
to further clean the glass.  I do this by coupling a magnetron from
a microwave oven through the viewport.  The entire chamber will
light up with a blue glow.  You should see the pressure rise a bit.
After several minutes of plasma cleaning, open the vacuum valve and
bring the chamber back down.

Valve off the vacuum to keep vaporized metal out of the vac system.
Apply current to the terminals so that the titanium wire heats
enough to have a significant vapor pressure.  This takes "a few"
(less than 12) volts at several tens of amps.  I use a large
unregulated DC power supply and a variac.  The proper temperature
must be achieved experimentally unless you have an optical pyrometer
(and used a sapphire window instead of glass.)  The brightness level
of a light bulb operating about 30% undervoltage is a good starting
place.

When the titanium is hot enough, it evaporates and the vapor travels
to all corners of the chamber.  It condenses to a thin film whenever
it touches a cool surface.  It coats all surfaces in the cooker but
most of the heavy vapor coats the glass right under the wire.
Again, experimenting is necessary to determine how much time is
needed for each deposition.

A dichroic coating is made up of up to 30 or so layers of thin films
separated by  dielectric (oxide) layers in between.  (Very greatly
simplified) The spacing is 1/4 wavelength of the color you want to
reflect.  In this simplified procedure, you're laying down metal
layers.  You have to repeat the vaporization procedure for each
layer you desire to lay down.  With titanium, only a few layers
yield interesting effects.

To achieve an oxide layer in this application, you can admit some
oxygen to the chamber and then establish another plasma discharge.
This will form atomic oxygen which oxidizes the surface of the
titanium film.  Again, much experimentation is required to perfect.

If you're really interested in this, I suggest that your first step
be to buy the book "Handbook of Thin Film Technology" ISBN,
0-07-039742-2.  Sucker cost me $104 back in '92.

Lastly,  Several years ago, I briefly chatted with a couple in
Atlanta who were doing Dichroic glass in their basement (or at least
had the equipment - not sure how well it was going.)  I have the
name of Dave & Sherry Moser in Norcross, Ga.  Sorry, don't have a
phone number.  If you want to do some digging, you might look them
up and have a chat.

John


 
Index Home About Blog