SEM vs EPMA

[lucataj]

Hello, i turn to you with a question which i have recently been getting because i want to buy a microprobe and the “money providers” several times questioned “why microprobe and not just high-resolution SEM”. My arguments are towards measuring the minor and trace elements. However, I was wondering if some of you did some real comparison (even major elements) or what is your experience or any hints how to argue for having the microprobe. i think this might be a general problem for future, if the grant agencies will consider microprobe as a “basic” “not necessary” “too expensive to purchase and to maintain compare to SEM”. thanks in advance for any hints! cheers lucie

[BenjaminWade]

Hi Lucie
It is a difficult question to answer. You need to identify what your primary driver is:

If it is just wanting to measure minor and trace elements then I would put money towards an excimer laser ablation unit with ICP-MS, which would be much cheaper than a microprobe.

If it is a lot of major/minor elements with sometimes trace elements, then I would get a microprobe.

If it is predominantly imaging with perhaps sometimes trying to quant major-minor elements then you could go perhaps go down the route of SEM with really good EDS and try some quant EDS with standards, or FEG-EPMA but sounds like the money wouldnt permit that. I am not sure what your samples might be, but many major elements can be quantified quite well these days with standardised EDS. There are still (and always are) exceptions to this though.

Obviously many trace elements can be done via EPMA with careful setup and analysis. However if spatial resolution isn't an issue, if it isn't metal/steel samples you want to quantify elements in, and if you have the standards, many low level elements are done much simpler with much better detection limits via LA-ICPMS. Bear in mind that quantification of LA-ICPMS data requires the input of the concentration of one major element that is present in your sample. Sometimes this can be stoichiometric, sometimes it has to be measured by another technique (such as EPMA...).

Cheers

[Probeman]

Hi Lucie,
This is such a great question.  I think we should be discussing questions like this more often in the scientific community.

Ben Wade makes excellent points: it really is a trade off between spatial resolution, sensitivity and accuracy. The answer to these questions depends entirely on the samples under consideration and the specific questions that need to be answered.

Can you describe in greater detail what samples you might be looking at and what your spatial resolution, sensitivity and accuracy characterization requirements might be?

I realize that such answers are not easy to provide as we never know exactly what will be walking into the lab at some point in the future, but currently what is your best guess as to which samples and what measurements you will need to be making?

Feel free to provide some specific examples of the measurement problems you think you will be encountering.
john