Date: Tue, 27 Sep 94 23:07:19 CDT
Subject: EDX window contamination
From: dfalcon@VNET.IBM.COM (by way of johnf@geology.wisc.edu (John Fournelle))

Condensation of oils on EDX detector windows to the point of
dripping and forming puddles (in the worst cases) as well as interfering
with light element analysis has been a problem for as long as EDX
detectors have been around!! I have successfully eliminated and
prevented the problem on SEM's and electron microprobes for twenty years
on numerous systems by installing a halogen microscope lightbulb (6V 20W)
into the system physically in front of and to the side of the detector
window. (Somewhere between one half inch to three inches away from the
window seems to work fine.) The bulb is powered with reduced voltage to
a moderately dim yellow brightness whenever the system is not used
(evenings and weekends). This keeps the window warm by radiation heating
so that the window is not a "getter" for everything in the system by
cryopumping when it is cold. This not only keeps the window clean, but
will safely clean off an oily window without risking physical damage.
(Of course, this light must be turned off when using the SEM and the EDX
detector. If the detector is sensitive to light, as most light element
capable detectors are, you may have to wait one to five minutes after
turning off the light for the detector to recover and be ready for x-ray
analysis.)
If you want to clean up the whole system and minimize contamination
buildup on samples and apertures, you can also install a dry nitrogen purge
through the chamber and through the pumping system which can also be used
evenings and weekends when the system is not used. This may involve turning
off diffusion or turbopumps as well as ion gauges and leaving valves open for
purging through the pumping system. This involves a delay in the morning
of ten minutes to half an hour to shut off the purge and get to normal
high vaccum after turning on pumps,etc.--but the results in system cleanliness
are well worth it.