Suggestions for trace-element level Ni, Co analysis of serpentine

[GarnetSnowball]

I would like to analyse serpentine group minerals for their Ni and Co content.

I'm fairly new to EPMA, so would like some advice on what I can do to ensure precise and accurate results.

Concentrations are probably well within the trace-element range (<0.1 wt%), so what would be the best way to approach this? Do we need to analyse the same emission line across multiple detectors, like is done with Ti-in-Qtz analyses?

Any suggestions or direction towards relevant material is appreciated!

The instrument is a JEOL8530F

[Probeman]

I would like to analyse serpentine group minerals for their Ni and Co content.

I'm fairly new to EPMA, so would like some advice on what I can do to ensure precise and accurate results.

Concentrations are probably well within the trace-element range (<0.1 wt%), so what would be the best way to approach this? Do we need to analyse the same emission line across multiple detectors, like is done with Ti-in-Qtz analyses?

Any suggestions or direction towards relevant material is appreciated!

The instrument is a JEOL8530F

It depends on what sensitivity is required.  For each doubling of photon counts, e.g., doubling the beam current or doubling the count time or 2 spectrometers vs. 1 spectrometer, one can expect the sensitivity to improve by the SQRT(2).

That said, remember the mantra: for trace elements the most important consideration is the background measurement (and any spectral interferences):

https://probesoftware.com/smf/index.php?topic=1535.msg12121#msg12121

For trace elements in serpentine minerals, I would do a careful background evaluation, check for interferences and probably utilize a blank correction using a Mg2SiO4 synthetic material:

https://probesoftware.com/smf/index.php?topic=1405.0
https://probesoftware.com/smf/index.php?topic=454.0

Of course for ultimate trace sensitivity in silicate minerals (moderate average Z), one should be using the MAN background correction. Literally twice the sensitivity in 1/2 the acquisition time:

https://probesoftware.com/smf/index.php?topic=307.0

See also Donovan et al., 2016 Amer Min.

[GarnetSnowball]

Thank you Probeman, who never sleeps, for your helpful advice.