How often do you repolish your standards?

[dawncruth]

Hi all,
I'm deciding the sample holders for our new probe and am leaning heavily toward the split stage where the standards hang out in the chamber for long periods of time. I recognize needing to recoat for different accelerating voltages, but that might necessitate having different standard sets with different coat thicknesses. I mean our bread and butter is 15 kV so those standards could be installed in the chamber making it harder to repolish/recoat the standards.  That said, how often do you repolish/recoat your standards? 


Dawn

[John Donovan]

Hi all,
I'm deciding the sample holders for our new probe and am leaning heavily toward the split stage where the standards hang out in the chamber for long periods of time. I recognize needing to recoat for different accelerating voltages, but that might necessitate having different standard sets with different coat thicknesses. I mean our bread and butter is 15 kV so those standards could be installed in the chamber making it harder to repolish/recoat the standards.  That said, how often do you repolish/recoat your standards? 

Hi Dawn,
I'm trying to understand why one would "recognize needing to recoat for different accelerating voltages"? 

You will definitely want to have *known* coating thicknesses on your standards and unknowns, if they aren't the same thickness.  Because as you know, the difference in coating thickness between ones standards and unknowns will most definitely affect quantification, particularly for very low energy emission lines (differential x-ray absorption by the coatings), e.g., O ka, and also for very low overvoltage x-rays (differential electron energy loss from the coatings), e.g., Fe Ka at 10 keV.

But I don't quite see why the coating thickness would need to be different for different electron beam energies. As long as the coating thicknesses are the same, or known and specified in Probe for EPMA as described in this topic, I think it shouldn't matter:

https://probesoftware.com/smf/index.php?topic=23.0

Using the Calculate Electron and X-ray Ranges dialog in CalcZAF, even a 5 keV beam penetrates almost 500 nm in carbon.  Are you thinking of looking at inclusions approaching 20 nm in size (your nominal coating thickness)?

5 keV, 2.26 grams/cm^3, Electron range radius = 0.437628 um

I'm probably missing something, so can you explain what you mean? 

As to when to re-polish ones standard mounts, the rule of thumb in my lab is: whenever the metals start looking a bit oxidized.  When they're not in the instrument, we keep our standards in a dessicator, which helps a little I think, but keeping them in the probe under vacuum should help more.  Though I sometimes wonder if someone should do a study on these assumptions.

I'm thinking of when I worked at Lawrence Berkeley Lab many years ago, on stable isotope sputter targets for their particle accelerators.  One of our targets for transuranic studies was 48Ca which was very expensive and would also oxidize quickly in air. At first we tried storing the targets in a vacuum cannister with a dessicant, but they still oxidized pretty quickly. So then we tried storing them under high pressure Ar (~200 psi) which seemed to really slowed down the oxidation of our 48Ca targets.  Our hypothesis was that under vacuum, the remaining oxygen (or H2O?) molecules had such long mean free paths that they all eventually got to strike the surface of the calcium targets and cause oxidation, but under high pressure Ar, the mean free path would be much shorter, and that would slow down the oxidation.  Well, that was the thinking anyway.

Here's a fun link:

https://www.sciencealert.com/this-calcium-isotope-is-so-rare-it-costs-500-000-for-just-2-grams

Back to EPMA, the other time we normally re-polish our standard mounts is when the carbonate or phosphate standards get too "chewed up" by the beam and we can't find any "clean" areas for standardization.

[dawncruth]

Hi John,
I read the workup by Pete McSwiggen (see attached section 9, figure 13) showing a decrease in x-ray intensity based on coating thickness.

Dawn

[Probeman]

Hi Dawn,
OK, I now see where you got the idea.

Yeah, he doesn't suggest using different electron energies for different coating thicknesses, but as I mentioned above, he states that the coating thickness affects the emission of high energy x-rays (in the case of Fe Ka) due to the drop in electron energy and therefore in low overvoltage situations, and for soft x-rays (in the case of Fe La) due to x-ray absorption by the coating.

And he goes on to say that therefore the unknown and standard need to have identical coating thicknesses (or as I said previously the different coating thicknesses need to be correctly specified in the correction software).

As Peter states in his conclusions:

Regardless of the methods used, care should be taken when carbon coating the samples to ensure that identical thickness are applied to both the standards and the unknowns, and the surfaces should be cleaned of any oxide layers or polishing residue.

I guess I would say that Peter and I are in complete agreement on this, I just added an additional alternative to simultaneous coating, that is if the coating thicknesses are different *and* also of known thicknesses, one can specify these different thicknesses in the correction procedure in Probe for EPMA.  Either way works fine.

I guess that's one good reason for polishing both the standards and unknowns, and coating them together, is that not only do you get similar coating thicknesses, but you are also get clean, (less) oxidized surfaces for ones metal standards. 

In practice I've found that our carbon evaporator coating thicknesses are quite reproducible using the method of red to blue color change on brass:

https://probesoftware.com/smf/index.php?topic=1252.0

That said, when I've needed to do careful measurements of oxygen at major level concentrations (where the standard dominates accuracy), I often re-polish and re-coat both the standards and unknowns together. But not always. We sometimes get uncoated metallurgical samples from some customers, and for those samples, we simply specify in PFE that the standards are coated with 20 nm of carbon and the unknowns are uncoated, and it seems to handle this nicely. Here is a detailed discussion on this very topic:

https://probesoftware.com/smf/index.php?topic=23.0

Hope this helps, and finally note that there's a typo in the pdf where Peter references fig 12, but I think he means fig 13 (as there is no fig 12!).

john

[dawncruth]

Hey John,
Thanks for the info!

Dawn

[Probeman]

Hey John,
Thanks for the info!

Dawn

Hi Dawn,
No worries.

I went into a little detail about using acrylic mounts to mount standard materials here:

https://probesoftware.com/smf/index.php?topic=172.msg8991#msg8991

These types of mounts are of course less "flexible" in terms of layout, than the individual sleeved standard mounts, but they are a heck of a lot easier to clean when that becomes necessary.

We've found that having half a dozen or so different acrylic mounts for the most common sample types, i.e., hydrous minerals, alkali glasses, ceramics, elemental metals, etc. covers almost all analytical situations.