Finding density from Continuum

[emma_fisi]

Hi Probe Hive-mind,

I have a question about obtaining density from probe/SEM data.

We've been doing large grids of analyses using Probe Software and the Thermo EDS system (Pathfinder), in order to try and determine the bulk composition of some meteorites. In order to get at the true bulk composition, we need to know the density of each phase, in addition to its chemistry. Whereas olivine/sulfide/metal grains are easy - the densities for different compositions have already been determined - the meteorite matrix is made up of clay-like stuff, and the density of this material is not known.

Does anyone out there know of any method that can take the EDS spectra that we have collected, and, using the chemistry, calculate the density? I have heard of methods of using the EDS background to determine density, but I am not sure of how to do this in practice.

As always, any help is greatly appreciated.
Em

[Probeman]

We've been doing large grids of analyses using Probe Software and the Thermo EDS system (Pathfinder), in order to try and determine the bulk composition of some meteorites. In order to get at the true bulk composition, we need to know the density of each phase, in addition to its chemistry. Whereas olivine/sulfide/metal grains are easy - the densities for different compositions have already been determined - the meteorite matrix is made up of clay-like stuff, and the density of this material is not known.

Does anyone out there know of any method that can take the EDS spectra that we have collected, and, using the chemistry, calculate the density? I have heard of methods of using the EDS background to determine density, but I am not sure of how to do this in practice.

As always, any help is greatly appreciated.
Em

Hi Emma,
I've been thinking about this for a few days and I'm drawing a blank on what physics would allow a determination of density at the micro scale.

You mentioned a method using EDS continuum spectra, but do you have any more info about this method?

Unless the sample was a film of known composition and thickness, I can't think of a way one might determine density.  Because if one knows the mass thickness and linear thickness, the density falls out nicely, but otherwise I can't think of anything.
john

[emma_fisi]

Hi John,

I have found a paper by Zanetta (see attached) that claims to be able to do such a thing, but to be honest, I am not sure I understand what they are doing. It looks to involve modeling of the continuum, combined with point analyses to check the chemistry of particular phases.

Cheers,
Em

[Nicholas Ritchie]

I'm skeptical.  One of the reasons that EPMA works is that density doesn't matter.  To extract density from the an x-ray spectrum, you need a property which is a function of density and not mass-density.   The only time this happens is when the z range is limited like in thin films or particles.   Otherwise the term in all the equations is rhoz and never rho.  So can you extract density information from particles and thin films (so long as you know there dimensions) but never from bulk.

So what is Zanetta seeing?  I believe that Zanetta is mistaking the dependence of the background intensity on mean atomic number with dependence on density.   There is a strong correlation between atomic number and density - low z materials  typically have low density and high z materials typically have high density (think carbon vs lead). 

I suspect that if he replotted the data in Figures 6 and 7 against mean atomic number the agreement would be every bit as good - probably better.

[Probeman]

I suspect that if he replotted the data in Figures 6 and 7 against mean atomic number the agreement would be every bit as good - probably better.

Especially if he uses the Z^0.7 fraction weighted average atomic number!   

https://probesoftware.com/smf/index.php?topic=1221.msg8475#msg8475