Presentations on EPMA

[John Donovan]

Attached below is the Powerpoint file for the talk I recently gave at UBC (Vancouver). The talk was hosted by Mati Raudsepp and entitled "New EPMA Methods to Improve Precision and Accuracy for Element Quantification".

Remember, you need to be logged in as a member to see attachments.  I'm happy to discuss any of the points covered.
john

[Probeman]

Here's a short PPT on why WDS EPMA is the ideal elemental tool for thin films and nano-particle characterization.  See attached (remember you must be logged in).

[Probeman]

Anette von der Handt and I are starting to put together some materials for explaining "best practices" in EPMA measurement science. 

But to start off, here's an example of "worst practices" in EPMA: using San Carlos olivine as a primary standard for Ni... where its concentration is (roughly!) 0.3 wt%!       Believe it or not there are actually a few labs out there that think this is a good idea.   Clearly they have not thought this through very well!

Here are a few schematics that I put together to try and explain what matters for major elements vs. trace elements with regards to standard selection. I would be very grateful for any comments and or suggestions to improve these explanations.  Here is the general schematic:



All this is an attempt to explain that the standard concentration, matrix correction and dead time (accuracies) are multiplicative corrections and therefore dominate at high concentrations, but whose effects scale down as the concentration decreases as shown here:



While the background modeling and interference correction accuracy dominates at trace level concentrations, but since they are subtractive in operation, they remain constant regardless of concentration and therefore are negligible at major concentrations as shown here:



Any comments or suggestions to improve this explanation?

[Probeman]

To summarize the slides in the previous post on "rules of thumb" standard selection and correction considerations for major vs. trace elements, here are two more slides that I would appreciate any comments on. First for major elements:



and here for trace elements:



Of course for minor elements, a bit of each applies to standard selection and the critical considerations, depending very much on the details...

As a follow up regarding standard sourcing, we all should work to obtain (see FIGMAS efforts) high purity single crystal end member synthetic materials with compositional accuracy defined by stoichiometry (ideally two materials for each common geological element in order to obtain a k-ratio).

And also perform the instrument calibrations described in the post linked below (and the several following posts) to discover exactly how well calibrated your instrument is for: dead time, picoammeter linearity and k-ratio consensus between your spectrometers (including your EDS detector) as described here:

https://probesoftware.com/smf/index.php?topic=1466.msg11102#msg11102

And if you have the Probe for EPMA software the entire procedure can be automated! First make sure your Probe for EPMA software is fully up to date (using the Help | Update Probe for EPMA menu), then just click on the Help | How to Use Constant K-ratios to Calibrate Your Instrument menu, and get step by step instructions as a pdf file.

[sem-geologist]

I just want to point out, that for trace composition due to first point "* High concentration, synthetic material... primary standard" - dead time correction is also important. It is of secondary importance to pulse counting of trace unknown - but still very important for standard measurement, where count rate will be significant and dead time correction is important.

[JonF]

As rules of thumb, they seem pretty good. But they do seem to be focused on what I refer to as the "traditional" WDS elements though e.g. Ka lines of 3d metals, Mg, Al, Si etc.

I agree that this is what most people are doing (whether or not this actually needs WDS is another matter!), but I find myself doing ever increasing amounts of light element work, where even for major concentrations of light elements, the interferences and background corrections are very important.

[Probeman]

I just want to point out, that for trace composition due to first point "* High concentration, synthetic material... primary standard" - dead time correction is also important. It is of secondary importance to pulse counting of trace unknown - but still very important for standard measurement, where count rate will be significant and dead time correction is important.

Yes, you are absolutely correct!

Which is why I should have added: assuming one runs their high concentration primary standard for the trace element at a relatively low beam current compared to your unknowns. Then dead time should not be an issue.

[Probeman]

As rules of thumb, they seem pretty good. But they do seem to be focused on what I refer to as the "traditional" WDS elements though e.g. Ka lines of 3d metals, Mg, Al, Si etc.

I agree that this is what most people are doing (whether or not this actually needs WDS is another matter!), but I find myself doing ever increasing amounts of light element work, where even for major concentrations of light elements, the interferences and background corrections are very important.

Yes, I was focusing on "rules of thumb" for the common geological elements. You are most certainly correct that for some light element (oxygen, nitrogen, etc.) situations, e.g., low sin thetas, background modeling and interferences can be significant corrections for light elements at any concentration.

Fortunately modern LDEs Bragg crystals tend to suppress higher order reflections, so it's not as bad as it used to be!   

[Probeman]

Here's a slide from my "worst practices of EPMA" presentation:



So exactly why is it such a bad idea to use San Carlos olivine as a primary standard for trace Ni (or Mn, or Zn or whatever)?  Here's the next slide:



In fact, based how how heterogeneous San Carlos olivine has been shown to be,

Fournelle, J. "An investigation of “San Carlos Olivine”: comparing USNM-distributed material with commercially available material." Microscopy and Microanalysis 17.Supplement 2 (2011): 842-843.

we shouldn't even be using it as a standard for anything!

[wrigke]

Here is a link from a Cameca EPMA webinar entitled, "Electron Probe Microanalysis of Irradiated Nuclear Materials: Challenges and Solutions"

https://www.youtube.com/watch?v=yo5aovLk6gI