Problems with Short Filament Life - And Some Common Solutions
Tungsten thermionic filaments in electron microscopes
One of the greatest frustrations to face any electron microscope operator is
when they experience short filament life and don't know why. The problem
itself is so frustrating because there can be so very many different
explanations for shortened filament life that it is virtually impossible to
know for sure just what might be the cause.
We would like to cite several of the most common reasons followed by a
discussion about how to solve the problem that is causing the short filament time:
Defective filaments:
Even famous microscope manufacturers with famous names, at times, have
suffered the experience of "bad batches" of filaments. We consider it
always wise to put into reserve one or two boxes of known "good" filaments
so that when you do start experiencing short filament life, you can
immediately, put in a known "good" filament and see if it too suffers the
fate of the other filaments or does it give the expected long life time.
Makers of new filaments are not quick to admit to having put out a bad batch
but it does happen and this is the "quick and dirty" way to determine
immediately whether a filament life-time problem is due to a bad batch of
filaments.
Impatience:
A tungsten thermionic filament should not be operated at a vacuum worse than
about 1 or 2 x 10-5. Hence, anyone turning on their high voltage
and obtaining beam current, even though the microscope lets you do that, is
inviting a shortened filament life. Typically, the safety (high voltage)
relays on most SEMs and TEMs permit the high voltage to be activated at a
vacuum still in the 10-4 so if you begin operating without waiting
for the vacuum to get down a bit better, you will experience filament lifetimes
less than would otherwise be possible.
Poor or dirty vacuum system:
For a diffusion pump system, and there is also the mechanical pump that "backs" the diffusion
pump, we have not one but two pumps to worry about and maintain. For example, if air has been
accidentally let into the system when the diffusion pump oil was hot, it is possible to literally
"crack" the oil, and cause it to decompose it into a liquid with significantly reduced properties.
The mechanical pump oil can also be degraded by a
variety of reasons, one being age (the oil should be changed once every six months or so), or
exposure to moisture or other agents that could degrade the oil.
Typically, clean oil is relatively clear (Santovac® 5 for example, is quite
clear, SPI Lobo Oil is a light brownish clear color) but when cracked it
literally starts to look dark brown or even black. When this happens, in
the case of the diffusion pump, it mush be completely disassembled and
cleaned before new oil is put in. In the case of the mechanical pump fluids,
a flush with some new fluid often times will return the pump to a clean
state, which then makes it ready for the charge of new fluid.
Disappearance of pump oils:
During the routine operation of any SEM or TEM, while it is not supposed to
happen, accidents do occur whereby air is partially sucked into the system,
which also causes some sucking out of the diffusion pump fluid into the
column and while the column can be cleaned, the pump is still missing some
of its charge of oil. In the case of the mechanical pump, the oil level
with time can drop and at times, oil has to be added. In the case of the
diffusion pump, however, it literally has to be dropped and inspected, and
if the remaining oil is still clear, then probably all that is needed is the
addition of more of the diffusion pump fluid.
Clogged oil mist filter:
The purpose of the oil mist filter is to capture the "mist" of oil that
would otherwise come out of the outlet of the mechanical pump. And with the
passage time, this filter does reach capacity and fill up and when that
happens, no only is it unable to catch more oil mist, but it causes a back
up of the entire vacuum system and with the lower pumping capacity, there is
less vacuum achieved. The solution of course is to replace the
oil mist filter.
Cooling water temperatures not within specification of the diffusion pump:
The diffusion pump has very specific specification for optimum operation,
and it is around those parameters that the vacuum systems of electron
microscopes are designed. The users most likely to experience problems of
this nature are those who are using cooling water from a tap and not via a
stand-alone water recirculator.
The inlet temperature must be controlled and the outlet temperature must be
controlled and therefore the flow rate must be adjusted to give the correct
exit temperature. If the exit temperature of the water is too high, then
the diffusion pump won't operate right and if the exhibit temperature is too
low, it won't operate as well either but with another downside: It will put
an extra strain on the diffusion pump heater coil which will result in early
failure of it as well.
Vacuum leaks around the filament housing:
Since the filament housing is what gets opened up each time a filament is
changed, unlike other "O" rings in the column, which are hardly ever
"opened" or "exposed", this one has multiple opportunities to collect dust,
get scratched, and generally "age", any of which of course could result in a
vacuum leak. If we suspect this is the problem, we recommend trying some
vacuum grease on that "O" ring, using either
Santovac® 5GB vacuum grease or the
Braycote® Micronic 803 vacuum grease.
Are the samples the culprits?
Some samples can literally be "off-gassing" either water or some organic
solvent, even when one is pretty sure it is "dry". So long as a sample is
off-gassing, good vacuum will be impossible to obtain. This problem is most
severe in humid climates. For example, a paper sample left out in the open
on a humid day could take 30-60 minutes to fully pump down to operating
vacuum. The problem again is that even when the vacuum gage says the vacuum
is good, there is still off-gassing and if those evolved species end up in
the filament area, there will again be a short lifetime situation.
To Ask a Question or Make a Comment
To Place an Order or Request a Quote
Tuesday May 13, 2008
© Copyright 2008. By Structure Probe, Inc.
Contacting SPI Supplies and Structure Probe, Inc.
All rights reserved.
All trademarks and trade names are the property of their respective owners.
Privacy Policy
Worldwide Distributors, Representatives, and Agents
