Adding non-traditional lines to PfE

[JohnF]

There were several presentations at M&M which discussed the utility of "non-traditional" x-ray lines in EPMA, particularly as there is increasing interest in the use of low keV EPMA (<10 keV down to ~5-6-7 keV). In particular the Fe LL (Ll, not L one) line. Adjacent to it is the LN (Ln, not L eta) line which is sometimes indistinguishable from LL. Gopon et al (2013) showed the usefulness of the LL line for the transition element Fe where the La and Lb lines are problematic in the correct EPMA of Fe-silicides, using Fe metal for the Fe standard.

I have also done some experiments at low keV for some REE compounds and similarly found M gamma and M zeta to be useful, as the Ma-Mb lines can be extremely user-unfriendly lines (Fischer-Baun J. Applied Physics Vol 38, p 4830+). Namely, it can be difficult to distinguish exactly where the Ma and Mb are, when one may be bifurcated by the edge. And then even if it can be correctly determined, the analytical results are extremely wrong (eg. in a SmMgZn alloy, the Sm was ~24% too high using Ma)--much like the problem using Fe La in Fe-silicides.

In attempting to utilize all of these 'non-traditional' lines in _acquisition_ there have been many problems. One must "trick" the acquisition software because all (*one exception recently) do not support acquisition of these lines, so one must choose as the analytical line (to be peaked up and faked out) such as Mb. The problem arises when one wishes to see the _correct_ display of adjacent x-ray lines, as the software displays all positions (angstroms, sin theta, mm, keV) relative to the 'central element' which say is listed as Sm Mb when in fact it is Sm Mz.

Recently Cameca implemented the ability to acquire many non-traditional lines in their Peak Sight software, a nice first step, and there remains followup work to implement the correct matrix correction of the lines.

"It would be nice" to be able to directly acquire (vs 'faking out') the software some of these non-traditional lines in PfE. But I understand this is a difficult thing to implement. But a goal to be aimed for.

[John Donovan]

"It would be nice" to be able to directly acquire (vs 'faking out') the software some of these non-traditional lines in PfE. But I understand this is a difficult thing to implement. But a goal to be aimed for.

Hi John,
It will be some work but I'm going to add it this summer.  I want to add six more analytical lines (for symmetry) and so far you and Phil have suggested Ll, Ln, Mz and Mg.

Please suggest two more lines, then let's tabulate emission and edge energies for them for elements 3 to 100 and I'll work on getting them imported in the PFE data files.

The MACs I'll calculate on the fly using McMaster (to  start with), but one can override this using the "empirical MACs" feature in PFE because I know these low energy MACs can be problematic.
john

[JohnF]

Very good. Here's some more details vis a vis Sm M lines, wavescans on the ORNL/USNM SmPO4 (15 keV, 60 nA, 8 secs/channel)

From that my recommendation is
1. MG Mgamma M3-N4,5 transition
in Nist database, this is 26% intensity of M max intensiy

2. MZ1 Mzeta1  M5-N3
in Nist database, this is 6% of M max intensity

Note that Mb>Ma in intensity in many/all? cases.

FYI, for the Sm phosphate std, I find both lines strong  (these are P-B intensities, 15kV):
MG is 23% of MB and
MZ is 26% of MB
—> and far from the M3 and M5 edges (major issue is the self absorption by the M5 edge for the Ma and Mb lines

Below are (1) wavescan showing Sm MG to the left edge, Sm MZ to the right edge, and the crazy mixed up MB-MA peaks in the middle, and (2) enlarged central area showing the M4 and M5 absorption edges in the middle region. Note that the M3 edge is far enough to left of MG and the M5 is "miles away" from the MZ peak.

[pgopon]

Does anyone else have any suggestions for non-traditional lines that should be added to PfEPMA? 
So far the list includes: Ll, Ln, Mg, Mz
Is there anything that we are forgetting in that list that might be useful (especially for the monazite/ High Z element analysis folks?)

phil

[John Donovan]

Does anyone else have any suggestions for non-traditional lines that should be added to PfEPMA? 
So far the list includes: Ll, Ln, Mg, Mz
Is there anything that we are forgetting in that list that might be useful (especially for the monazite/ High Z element analysis folks?)

Hi Phil,
Nice hanging with you at M&M!

Karsten (geoman!) Goemann and I are are having a beer and here is the list we came up with:

Ll   (L-III edge)   L3-M1
Ln   (L-II)         L2-M1
Lv   (L-II)         L2-N6(7)
Ly   (L-II)         L2-N4
Mg   (M-III)        M3-N5
Mz   (M-V)          M4,5-N2,3


How does that sound?

I just also realized that I already have the edge energies for all these, but I still need the emission energies for elements 3 to 100.  What about fluorescent yields?
john

[Mike Jercinovic]

John, I have a pretty comprehensive table of emission lines with energies and wavelengths if you want it.

[John Donovan]

John, I have a pretty comprehensive table of emission lines with energies and wavelengths if you want it.

Hi Mike,
Can you extract just the lines described above to an Excel or ASCII file?   If these are the lines we have decided on...

Also, is there a preferred order to the new emission lines?  They will (of course) be listed after the Ka, Kb, La, Lb, Ma, Mb lines, but after that...
john

[Mike Jercinovic]

I will send an Excel file.  If you just want emission energies, sort by order, then cut the high order lines, then sort by line.  As for how to list, I guess one of two ways, either by energy for an element, or element-shell-predicted intensity (so always starting with the most intense line for a particular shell). Maybe others will have a strong opinion about this.

[Philipp Poeml]

I could and try to provide some lines for the actinides, experimental values, but usually much better than the tabulated values. However, not that e.g. the Np Mg line is split into two lines, probably a Np Mg and a resonant line. It is not always so easy at high z... ;-)

[John Donovan]

I could and try to provide some lines for the actinides, experimental values, but usually much better than the tabulated values. However, not that e.g. the Np Mg line is split into two lines, probably a Np Mg and a resonant line. It is not always so easy at high z... ;-)

Hi Phil,
Indeed. Why do you think I asked you to do it for us!   

I'm wondering though: should we replace the Lv with the Lg line?  Which is stronger intensity?  John F. asked about the Lg line.
john

[Mike Jercinovic]

Lg1 (L2-N4) is, in principle, 1000x more intense than the Lv line (L2-N6,7).

[John Donovan]

Lg1 (L2-N4) is, in principle, 1000x more intense than the Lv line (L2-N6,7).

Oops, I guess I already had Lg (gamma).

So should we use something else other than Lv?

Ll   (L-III edge)   L3-M1
Ln   (L-II)         L2-M1
Lv   (L-II)         L2-N6(7)
Ly   (L-II)         L2-N4
Mg   (M-III)        M3-N5
Mz   (M-V)          M4,5-N2,3

[Mike Jercinovic]

Maybe the NIST X-ray database is the best source of complete energies (?)
http://physics.nist.gov/PhysRefData/XrayTrans/Html/search.html

It is certainly a quick way to list out all the energies (that is, all elements) for a particular transition.  And to evaluate theoretical vs measured values (of course, depending on how you measure and in what material!)

[Philipp Poeml]

This NIST table seems to be free of M lines...

[John Donovan]

This NIST table seems to be free of M lines...

The NIST based x-ray table (XRAY.MDB) that comes with the CalcZAF distribution includes K, L, and M lines and edge energies for K, L, M, N and O shells.

http://probesoftware.com/smf/index.php?topic=286.msg2600#msg2600

The input file for this is attached below (you must be logged in to see it).
john