Carbon PHA needing SD-WDS!

[Ben Buse]

Hi.

I was measuring carbon in matterials containing oxygen, and the need for better PHA resolution becomes obvious.

PHA scan difficult to seperate from electronic noise



The poor resolution of the voltage peak - its very wide - making it difficult to cut out 2nd order interferences



Here's a wavescan for minerals with oxygen.



Fortunetly the 2nd order oxygen peak is very narrow and does not seem to contribute to the intensity under the C peak. But might affect measurment of very low concentrations such as in glasses. It be interesting to see what a WDS with a solid state detector instead of proportional counter would do --  R Wuhrer1 and K Moran2
http://iopscience.iop.org/article/10.1088/1757-899X/304/1/012021

Ben

[BenjaminWade]

Hi Ben
I would be very interested to see how you go measuring carbon (in some minerals I presume?). I tried (and failed!) this year to do some lower level C measurement in Scapolite (~0.6wt% C). My problem wasn't necessarily the O peak, but damn carbon on everything. I repolished a second standard mount I had and Ag coated it instead, which in the end was probably not a good idea due to a potential Ag intf on C position. But as you see from your wavescans even on pure Ag metal its a small peak, and if you use a 5nm coat it was my hope it would be insignificant. I investigated blank subtraction but the problem was that the "background" C signal was not consistent across my standards... so I put it in the too hard basket for the time being.



In the scans above 501 and 535 are both feldspars which I was assuming shouldn't have measurable carbon in them, the red scan is the Smithsonian scapolite (~0.6 wt% C).
Need an anticontaminator I guess while analysing for lower level work. I think John had a separate thread somewhere looking into that.

P.S. Ignore the fact that my peak is way off.....at the time my PC0 was in desperate need of alignment...the dynamic offset was something ridiculous.

[Probeman]

Need an anticontaminator I guess while analysing for lower level work. I think John had a separate thread somewhere looking into that.

Hi Ben,
Trace carbon is difficult in an EPMA, but it can be done.  Did you try the TDI correction?

It helps to have a cryo pump system in your EPMA:

https://probesoftware.com/smf/index.php?topic=646.msg3823#msg3823

Of course you have to coat non-conductive samples with something other than carbon.  As you mentioned there is a Ag interference on C Ka, but I've used an Al coating at times and it seems to work well for trace carbon analyses, but the unknowns and standard really ought to be coated together, unless of course one specifies different coatings in PFE for the standards and unknowns.

Here is a long post that discusses several issues for trace carbon:

https://probesoftware.com/smf/index.php?topic=1005.msg6659#msg6659

[BenjaminWade]

Hi John
Yes I did try the TDI at the time to see if I could get more consistency, but no dice.
I also looked into Al at the time (and I remember you discussing it), but I had no avenues of doing that either with evap or sputter. I did try coating both samples and standards at same time with Ag as well. Part of the problem might have been that I could only sputter coat the Ag which probably gave me a lot of islands.

I have an oxygen air-jet setup on the old SX51 which is not operational, but have no experience with it and/or how difficult that would be to install on the SXFive. I know people have had success with nitrogen trap/cold finger, stage heating, airlock plasma cleaner as well. I don't have anywhere near enough demand for the work to invest in them though.

Cheers

[Probeman]

Yes I did try the TDI at the time to see if I could get more consistency, but no dice.

Interesting.  I suspect it's because you have heavy carbon contamination.  And carbon contamination can be very inconsistent.

I also looked into Al at the time (and I remember you discussing it), but I had no avenues of doing that either with evap or sputter. I did try coating both samples and standards at same time with Ag as well. Part of the problem might have been that I could only sputter coat the Ag which probably gave me a lot of islands.

That is correct. Sputter coating is simply unsuitable for quantitative work. An evaporator is essential.

I have an oxygen air-jet setup on the old SX51 which is not operational, but have no experience with it and/or how difficult that would be to install on the SXFive. I know people have had success with nitrogen trap/cold finger, stage heating, airlock plasma cleaner as well. I don't have anywhere near enough demand for the work to invest in them though.

Yes. You'll need a relatively clean system in order to attempt trace carbon analyses. Remember, necessity is a mutha!