X-ray Mapping: No ROI outputs for "new" Thermo NSS? A step backward?

[JohnF]

Recall that we are installing a new SXFive FE probe here currently, and last week while I was at AGU, Thermo FSEs came to install the NSS. Apparently, I hadn't asked the right questions of the right people previously, and now have since learned that what some might call "advanced" features have eliminated some truly useful features that the "old" EDS systems had.

Following discussions with Julien at AGU and then looking thru the Forum, I have come to a realization that what I had "assumed" to be possible with the new probe, is apparently impossible. I had "assumed" that I could do all my mapping, particularly quant with MAN, in Probe Image, having "full" access to BOTH 5 WDS channels and as many EDS ones as I wanted (as I had with MicroImage). Now I learn that I cannot do that, as Thermo doesn't have the ROI outputs.

Yes, I can have Thermo upgrade me to input all 5 WDS channels into their software but to me that is a poor second to mapping with PI. I don't want NSS's limited/rigid/fixed size options for mapping area and I don't want to have to rescan over, etc, etc. Yes, I know Thermo can output their jpg x-ray maps (like they apparently can do with Ken's Jeol), but that is a poor second to having x-ray counts per pixel. And does nothing for MAN quant mapping.

A key problem is that I cannot have the elemental ROI counts for each pixel and then run the maps thru MAN to make fully quant images. Thus, whereas I should have been able to run on ONE pass 5 WDS channels for minor elements and say 6-7 EDS channels, now I have to run TWO passes over the same region to get only 10 elements, for MAN maps.

So as far as anyone know,  is  there no way to get discrete EDS channels out of thermo into PI? And this isn't something Thermo has any interest or abiilty to change?

[Probeman]

So as far as anyone know,  is  there no way to get discrete EDS channels out of thermo into PI? And this isn't something Thermo has any interest or abiilty [sic] to change?

I know the Thermo hardware used to have this EDS ROI output capability in their previous Voyager product, so the possibility exists... Have you asked Thermo about it?

But even if you had the hardware EDS ROI outputs as does Bruker, they still need to be background corrected and quantified. For example, you mention the MAN background correction for EDS, but how would that work?  I think a better solution would be to use MAN backgrounds for WDS elements and... then what for EDS?

[John Donovan]

A key problem is that I cannot have the elemental ROI counts for each pixel and then run the maps thru MAN to make fully quant images. Thus, whereas I should have been able to run on ONE pass 5 WDS channels for minor elements and say 6-7 EDS channels, now I have to run TWO passes over the same region to get only 10 elements, for MAN maps.

Hi John,
This is a wonderful issue and warrants an extended discussion. I will start a new topic in the General EDS board because while Bruker does have EDS ROIs output for mapping (Thermo does not as you point out), the question of EDS-WDS quantification, especially if one wants to utilize the EDS ROIs for mapping major elements, is still problematic.

Note that the issue of integrated EDS and WDS quant for full quantitative point analyses is fully solved in Probe for EPMA as shown here:

http://probesoftware.com/smf/index.php?topic=79.0

But fully quantitative x-ray mapping with both EDS and WDS is still problematic for everyone I think, though there are some partial solutions, for example, the "hyper-imaging" cursor feature (which still needs to quant the EDS spectrum image pixels) as seen here:

http://probesoftware.com/smf/index.php?topic=83.0

[JohnF]

So as far as anyone know,  is  there no way to get discrete EDS channels out of thermo into PI? And this isn't something Thermo has any interest or abiilty [sic] to change?

I know the Thermo hardware used to have this EDS ROI output capability in their previous Voyager product, so the possibility exists... Have you asked Thermo about it?

But even if you had the hardware EDS ROI outputs as does Bruker, they still need to be background corrected and quantified. For example, you mention the MAN background correction for EDS, but how would that work?  I think a better solution would be to use MAN backgrounds for WDS elements and... then what for EDS?

Why wouldn't collecting EDS backgrounds work? e.g. using MgO for the Fe MAN background? Of course, one would have to do a bit of empirical research to find where there are no peak or shoulder overlaps, but in theory it should work... so be counting on EDS channels whilst collecting WDS MAN bkgs...

[John Donovan]

Why wouldn't collecting EDS backgrounds work? e.g. using MgO for the Fe MAN background? Of course, one would have to do a bit of empirical research to find where there are no peak or shoulder overlaps, but in theory it should work... so be counting on EDS channels whilst collecting WDS MAN bkgs...

Yeah, but PFE acquires full EDS spectra, not EDS ROIs. So it's apples and oranges...

[Probeman]

Why wouldn't collecting EDS backgrounds work? e.g. using MgO for the Fe MAN background? Of course, one would have to do a bit of empirical research to find where there are no peak or shoulder overlaps, but in theory it should work... so be counting on EDS channels whilst collecting WDS MAN bkgs...

Yeah, but PFE acquires full EDS spectra, not EDS ROIs. So it's apples and oranges...

I started a new topic here on this general issue here:

http://probesoftware.com/smf/index.php?topic=400.0