Danny get back to me for full details.  We manufacture surface tension
modification equipment.  We have recently built a unit to alter the surface
tension on the walls of a capillary tube manufactured like your but this was a
glass base with a glass etched tube but in the low microns cross section but
hundreds of microns long.  The complete answer for Teflon adhesion is a plasma
treatment.  I am in the middle of testing Teflon modification by plasma for a
number of customers.  The tool of choice to measure the Teflon reaction is a
goniometer.  This measures the contact angle of water, or any liquid, with the
substrate surface.  Teflon is very hydrophobic the contact angle of water is
above 100 degrees.  This means water or any liquid glue will bead up on the
surface and roll off if allowed to.  Using our flat plasma reactor to allow
constant plasma intensity we achieved the following :- using a 1 minute Argon
plasma the contact angle was down to less than 3 degrees.  Total hydrophilicity
and any liquid will stick to the Teflon like glue, excuse the pun.  After a week
of sitting in the open waiting for adverse reaction to set in the contact angle
is just about 5 degrees still very hydrophillic.  Using a shorter plasma time 10
seconds in argon plasma we get about a 45 degree contact angle.  We are still
reading 45 degrees, 72 hours later.  Both of these samples are left out in
normal room air.  Obviously you can do a plasma for just the right amount of
time and glue within a reasonable time period.  this is the only method I know
to safely glue Teflon.  Where are you located geographically, if close to San
Jose you could visit and try samples.  If not mail may work.

Bill Moffat
Yield Engineering Systems