Limits to X-ray map quantification: accuracy vs. precision

[Probeman]

Just finished the Athens EPMA Workshop organized by Ery Hughes and Ioannis Baziotis and it was fantastic. Job well done everyone.

One more thought I had with regards to one of the questions from Radek Michallik (I think), was regarding limits of X-ray map quantification regarding accuracy and precision.

First I would refer those interested to the pre-print of our quant mapping paper that will be published fairly soon in American Mineralogist:

http://www.minsocam.org/MSA/Ammin/AM_Preprints/7739DonovanPreprint.pdf

But I wanted to add that although accuracy of quantitative X-ray maps using full deadtime, beam current, background and matrix corrections is expected to be the same as point analyses (of course precision will certainly be less unless one utilizes pixel averaging), the accuracy of X-ray mapping at very short pixel dwell times could be affected by the accuracy of the scan generator.

That is, if the scan generator is supposed to produce 20 msec pixel dwell times, but actually produces a 19 msec dwell time, there will be a 1/20 or 5% accuracy error due to this electronics miscalibration. 

I must say say that I haven't seen significant problems with map quantification using 100 to 500 msec pixel dwell times, but I haven't tested accuracy at shorter dwell times. However the precision might not be sufficient at such short dwell times to evaluate accuracy.

Usually I utilize several hundred msec dwell times because want to attain decent sensitivity without too much beam current, but if one is utilizing very high beam currents, one might get enough precision at very short dwell times, but then deadtime effects might be also a problem...

It would be interesting to hear form other labs on what they have observed with regards to pixel dwell time accuracy at short dwell times.

[Probeman]

The full paper on quantitative mapping is now available here:

https://epmalab.uoregon.edu/pdfs/Donovan_2021_Amer_Min_2021-7739.pdf