JEOL Probe Users Listserver
Moderator: Ellery Frahm, [log in to unmask],
Electron Microprobe Lab, University of Minnesota
Post a message: send your message to [log in to unmask]
Unsubscribe: send "SIGNOFF PROBEUSERS" to [log in to unmask]
On-line help and FAQ: http://probelab.geo.umn.edu/listserver.html
Charging can be subtle as you and others have mentioned. Especially
if using low overvoltage conditions, e.g., Fe ka at 10 keV. This
could easily be a contributing factor in Ellery's low totals.
Regarding MACs, I don't think this is likely to produce systematic
errors large enough to explain Ellery's totals, but it is possible.
For silicates/glasses the worst problem I know of is that the 1966
(Heinrich compiled table) MAC for Mg ka in Fe is about 20% too high.
This is the culprit responsible for so many garnet and olivine totals
being too high that some may have noticed. However this problem only
contributed about 1% absolute to totals in worst cases and even then
the error is in the wrong direction for low totals.
Here are some comparisons:
MAC value for Mg ka in Fe = 5239.398 (LINEMU.DAT) Henke 1988
MAC value for Mg ka in Fe = 6120.7 (CITZMU.DAT) Heinrich 1966
MAC value for Mg ka in Fe = 5395.104 (MCMASTER.DAT) Livermore
MAC value for Mg ka in Fe
= 5518.652 (MAC30.DAT) Heinrich calculation
MAC value for Mg ka in Fe = 6089.534 (MACJTA.DAT) Frame equations
MAC value for Mg ka in Fe = 5093.8 (FFAST.DAT) NIST
Note the recent FFAST value from NIST is even lower than the 1988
Henke value. FFAST is probably the most accurate table available
today, but I would still use empirically measured MACS for oxygen,
carbon etc from the literature (Bastin, Pouchou, et. al.).
The guy who can say more about this is Paul Carpenter but he is
teaching at the Lehigh short course this week.
At 08:00 AM 6/12/2008, you wrote:
>JEOL Probe Users Listserver
>Thanks for your comments.
>>I have had specimens that look OK and don't appear to charge but
>>gave low results.
>>When I used my EDS/EDX detector and checked where the backgound x-
>>rays were going to zero counts at the high energy end (the Duane-
>>Hunt limit), I found that I had about 2kV of charge build up that
>>was decelerating the beam. The standards did not have this charging
>>problem. So I was effectively measuring standards at one voltage and
>>doing analysis at another. On my machine it's easy to do a
>>qualitative ED analysis while the probe is doing a WDS analysis.
>>Coating grounds the surface of the specimen but does not always
>>remove sub-surface charge build up.
>Checking the DH limit is an excellent idea -- we'll have to check that
>>This has also caused me trouble before. My old 8800 JEOL (15 years)
>>uses Henke (1966) MACs for longer wavelenghts ( > 1.2nm) and
>>Heinrich (1966) MACs for shorter wavelenght x-rays. My newer 8200
>>EPMA (4 years) claims to use Henke (1966, 1982) for the longer
>>wavelenghts and the same Heinrich (1966) MACs for shorter wavelenght
>>x-rays. When I looked up the actual value in my new probe for O Ka
>>in Si I was given 8790.0, the same value John gave before as Bastin
>>1992. I don't know the age of your probe, or what MACs it uses, but
>>changing MACs has affected my results (different sample) by about
>>the same amount you are talking about. You may be able to edit you
>>MAC table and recalculate to see if it makes a difference for you.
>We have an 8900, but we also have the latest software updates, so I
>don't know if it still uses the original 8800 files or the newest 8200
>files. I'm not sure if this could explain our low totals even when
>we're not measuring oxygen, but that's a very good question! (Hint,
>hint: any JEOL folks want to chime in either on- or off-list?)