Wish List for PFE Features

[Ben Buse]

PHA Scan peak fit

After peaking an element we often acquire the PHA bias scan. However, in the output (Display PHA, Peaking and Peak Scan Data) we get a peak fit only for the ROM Scan for the peak position, which is automatically updated for the element in the run.

How difficult would it be to implement a peak fit for the Bias Scan, where the peak is not always as smooth as a spectrometer position peak scan?

Or would it be at least possible to implement some feature where, similar to the peak scan position update, upon clicking in the graph, the new value for the bias is automatically updated for the element (would need to be an integer without decimals) ?



Until now we need to determine the new bias by hand from the graph and note the value on a piece of paper (or the more advanced of us use excel or other digital media  ), then go to the PHA dialogue and type it in for every element. Now this might be not such a big deal for the daily routine analysis, but when the routine becomes a set of 20 and more elements it is quite time consuming to do this manually. Sorry for being spoiled...

Rgds,
Radek

This was a fantastic suggestion Radek.

Thanks for implementing this John. Just used it for the first time in version 13.8.3, and it turned one of my least favorite method setup chores into a 90 second breeze!

Agreed click on bias position to input new value a great implementation, thank you John

[John Donovan]

PHA Scan peak fit

After peaking an element we often acquire the PHA bias scan. However, in the output (Display PHA, Peaking and Peak Scan Data) we get a peak fit only for the ROM Scan for the peak position, which is automatically updated for the element in the run.

How difficult would it be to implement a peak fit for the Bias Scan, where the peak is not always as smooth as a spectrometer position peak scan?

Or would it be at least possible to implement some feature where, similar to the peak scan position update, upon clicking in the graph, the new value for the bias is automatically updated for the element (would need to be an integer without decimals) ?



Until now we need to determine the new bias by hand from the graph and note the value on a piece of paper (or the more advanced of us use excel or other digital media  ), then go to the PHA dialogue and type it in for every element. Now this might be not such a big deal for the daily routine analysis, but when the routine becomes a set of 20 and more elements it is quite time consuming to do this manually. Sorry for being spoiled...

Rgds,
Radek

This was a fantastic suggestion Radek.

Thanks for implementing this John. Just used it for the first time in version 13.8.3, and it turned one of my least favorite method setup chores into a 90 second breeze!

Agreed click on bias position to input new value a great implementation, thank you John

Thank-you very much Ben!  Feedback is always appreciated!

Yes, sometimes it's the small "tweaks" to the software that can have a very positive impact on the day to day operation of EPMA instruments.

To everyone: in fact, it's worth taking a look back through this topic to see what little things have been requested/implemented by/for users over the years.  For example this tweak that allows one to export sample positions from Probe for EPMA in the PictureSnapApp annotation format for importing into PictureSnapApp:

https://smf.probesoftware.com/index.php?topic=71.msg12528#msg12528

This allows the full power of PictureSnapApp annotation options to be utilized for analyses from PFE. Here's a webinar on PictureSnap and PictureSnapApp on our YouTube channel:

https://www.youtube.com/watch?v=Q6tzYdyOJmE&ab_channel=ProbeSoftwareInc

And here are some examples using PictureSnapApp:

https://smf.probesoftware.com/index.php?topic=1020.msg6876#msg6876
https://smf.probesoftware.com/index.php?topic=1082.msg7340#msg7340

[Ben Buse]

Don't know if this of wide interest, but it would be good to load in a file setup - without the peak position, pha or bias. i.e. load in the count times, backgrounds and elements, but otherwise use the information in the current file. For when calibrated all elements for two setups, and then want to load in the two setups from another file.

[John Donovan]

Don't know if this of wide interest, but it would be good to load in a file setup - without the peak position, pha or bias. i.e. load in the count times, backgrounds and elements, but otherwise use the information in the current file. For when calibrated all elements for two setups, and then want to load in the two setups from another file.

It would be complicated because the current sample might be partially or completely different than the file setup being loaded.

Why not just load the two file setups one at a time, or even better (which is what I do): that is if you're going to have two sample setups with some elements in common (Si,  Fe, Ca), say olivine and feldspar, but will also have some elements that are not shared between the two setups (Cr, Na), create a "mother of all elements" sample with all the elements, tune that up, with the correct peak positions, PHA bias, etc., save it as a sample setup ("mother of all" sample setup). Then split it up into two different sample by making first, an olivine sample from the "mother of all" sample, then delete the elements you don't need for olivine, then again make a new sample (again using the "mother of all" sample, and again delete the elements you don't need for feldspar.

If you want to do that again in a new run, load the mother of all elements sample from that run using Load File Setup, then re-tune or not, then split it up again.

Or am I missing something?  What is your goal here?

[Ben Buse]

s, tune that up, with the correct peak positions, PHA bias, etc., save it as a sample setup ("mother of all" sample setup). Then split it up into two different sample by making first, an olivine sample from the "mother of all" sample, then delete the elements you don't need for olivine, then again make a new sample (again using the "mother of all" sample, and again delete the elements you don't need for feldspar.

If you want to do that again in a new run, load the mother of all elements sample from that run using Load File Setup, then re-tune or not, then split it up again.

Hi John, I guess I'm trying to avoid having to split it each time... but I agree it's complicated - it would have to check if element was present in current setup and correct spectrometer - and if so load apply parameters... so probably not feasible... it was just a thought... could I save time... and make sure I do the same split as previously

[John Donovan]

s, tune that up, with the correct peak positions, PHA bias, etc., save it as a sample setup ("mother of all" sample setup). Then split it up into two different sample by making first, an olivine sample from the "mother of all" sample, then delete the elements you don't need for olivine, then again make a new sample (again using the "mother of all" sample, and again delete the elements you don't need for feldspar.

If you want to do that again in a new run, load the mother of all elements sample from that run using Load File Setup, then re-tune or not, then split it up again.

Hi John, I guess I'm trying to avoid having to split it each time... but I agree it's complicated - it would have to check if element was present in current setup and correct spectrometer - and if so load apply parameters... so probably not feasible... it was just a thought... could I save time... and make sure I do the same split as previously

Or you could import each sample setup, peak it and then import then next sample setup, peak it, etc.

I prefer loading a file setup containing the "mother of all setups" so I can just run the re-standardization using that sample setup.

[Ben Buse]

Thanks John

Another thinking aloud,

Would anyone else be interested in calculating hydrous phases e.g. chlorite on an anhydrous basis, and would this be possible

Select h by difference, to ensure cation calculations have correct matrix correction; but then ignore H2O when calculating atoms per formula unit

[John Donovan]

Thanks John

Another thinking aloud,

Would anyone else be interested in calculating hydrous phases e.g. chlorite on an anhydrous basis, and would this be possible

Select h by difference, to ensure cation calculations have correct matrix correction; but then ignore H2O when calculating atoms per formula unit

I'm certainly no mineralogist, but the formula for chlorite is (Mg,Fe)3 (Si,Al)4 O10 (OH)2 · (Mg,Fe)3 (OH)6, correct? That simplifies to (Mg,Fe)6 (Si,Al)4 O10 and (OH)6.  Did I do that right? Typing that formula, (Mg,Fe)6 (Si,Al)4 O10 (OH)6, into the Standard app using formula entry (assuming 1:1 Mg to Fe and 1:1 Si to Al) we obtain:

St  250 Chorite syn
TakeOff = 40.0  KiloVolt = 15.0  Density =  5.000

for simulation
Oxide and Elemental Composition

Average Total Oxygen:      26.577     Average Total Weight%:  99.705
Average Calculated Oxygen:  48.466    Average Atomic Number:  16.128
Average Excess Oxygen:    -21.889     Average Atomic Weight:  24.514

ELEM:     FeO    SiO2   Al2O3     MgO       O     HO
XRAY:      ka      ka      ka      ka      ka        
OXWT:  44.754  24.952  21.171  25.106 -21.889  5.611
ELWT:  34.787  11.663  11.205  15.140  26.577   .333
ATWT:  55.847  28.086  26.982  24.305  16.000  1.008
KFAC:   .3081   .0796   .0682   .0893   .1466  .0000
ZCOR:  1.1290  1.4644  1.6425  1.6946  1.8129 1.0000
AT% :  15.315  10.210  10.210  15.315  40.840  8.111
24 O:   9.000   6.000   6.000   9.000  24.000  4.767

But the oxygen (oxide) weight percent is quite negative. Specifying Fe as Fe2O3 makes this worse. Can a mineralogist explain this to me?

[AndrewLocock]

Hello,
No, that is not the correct formula for chlorite.

There are currently 9 species in the chlorite group, and 2 related species, as follows:

Chlorite species   
baileychlore   [Zn5Al](Si3Al)O10(OH)8
chamosite   [Fe5Al](Si3Al)O10(OH)8
clinochlore   [Mg5Al](Si3Al)O10(OH)8
nimite   [Ni5Al](Si3Al)O10(OH)8
pennantite   [Mn5Al](Si3Al)O10(OH)8
   
donbassite   [Al4.33](Si3Al)O10(OH)8
   
borocookeite   [LiAl4](Si3B)O10(OH)8
cookeite   [LiAl4](Si3Al)O10(OH)8
sudoite   [Mg2Al3](Si3Al)O10(OH)8
   
Related species    
glagolevite   Na[Mg6](Si3Al)O10(OH)8·H2O
franklinfurnaceite   Ca2[Mn4Fe](Si2Zn2)O10(OH)8

[AndrewLocock]

Hello,
Continuing on chlorite.

I have an Excel spreadsheet at: https://www.eas.ualberta.ca/eml/files/Chlorite_spreadsheet2015.xlsx
It calculates both hydroxyl content and estimates ferric- and ferrous-iron proportions.
However, its assumptions for ferric iron calculation are only truly correct for trioctahedral chlorite (e.g., baileychlore through pennantite).
This spreadsheet does not deal with dehydrogenation (oxygen in place of hydroxyl).

Example analyses are included from volume 3B of the 2009 version of Deer, Howie and Zussman's Rock-Forming Minerals.

In Probe-for-EPMA software, I usually assume 8 hydrogen for 18 oxygen (H:O = 0.44444).
I rarely bother to try and calculate ferric iron in chlorite because of the large uncertainty in charge balance that results from error propagation of the measured elements.

You may also wish to look at:
A Windows program for chlorite calculation and classification
Computers & Geosciences 81, 101-113 (2015)
https://www.sciencedirect.com/science/article/pii/S009830041500103X

I hope that this proves useful.
Cheers, Andrew

[John Donovan]

Hello,
No, that is not the correct formula for chlorite.

There are currently 9 species in the chlorite group, and 2 related species, as follows:

Chlorite species   
baileychlore    [Zn5Al](Si3Al)O10(OH)8
chamosite    [Fe5Al](Si3Al)O10(OH)8
clinochlore    [Mg5Al](Si3Al)O10(OH)8
nimite    [Ni5Al](Si3Al)O10(OH)8
pennantite    [Mn5Al](Si3Al)O10(OH)8
   
donbassite    [Al4.33](Si3Al)O10(OH)8
   
borocookeite    [LiAl4](Si3B)O10(OH)8
cookeite    [LiAl4](Si3Al)O10(OH)8
sudoite    [Mg2Al3](Si3Al)O10(OH)8
   
Related species   
glagolevite    Na[Mg6](Si3Al)O10(OH)8·H2O
franklinfurnaceite    Ca2[Mn4Fe](Si2Zn2)O10(OH)8

Thanks!

OK, I tried your Clinochlore formula in the Standard application formula entry by replacing the square brackets with parentheses:  [Mg5Al](Si3Al)O10(OH)8  ->  (Mg5Al)(Si3Al)O10(OH)8

and now I get:

St  250 Chlorite (ideal clinochlore)
TakeOff = 40.0  KiloVolt = 15.0  Density =  5.000

[Mg5Al](Si3Al)O10(OH)8
Oxide and Elemental Composition

Average Total Oxygen:       51.816     Average Total Weight%:  100.000
Average Calculated Oxygen:  51.816     Average Atomic Number:   10.168
Average Excess Oxygen:        .000     Average Atomic Weight:   15.439

ELEM:     SiO2   Al2O3     MgO       O     H2O
XRAY:      ka      ka      ka      ka         
OXWT:   32.432  18.345  36.258    .000  12.966
ELWT:   15.159   9.709  21.864  51.816   1.451
ATWT:   28.086  26.982  24.305  16.000   1.008
KFAC:    .1063   .0623   .1575   .2993   .0000
ZCOR:   1.4256  1.5576  1.3879  1.7312  1.0000
AT% :    8.333   5.556  13.889  50.000  22.222
24 O:    4.000   2.667   6.667  24.000  10.667

Is this correct?  (Andrew says yes!)

[Probeman]

Ok based on Andrew's suggestions I decided to try doing a chlorite analyses in simulation mode in Probe for EPMA using the clinochlore composition from Andrew's above post.

I entered the clinochlore formula into the Standard database using the Enter Atom Formula Composition button and pasting in Andrew's formula and changed the square brackets to parentheses. Then I started a new probe run analyzing for Si, Al, Mg and Fe using SRM K-412 as the primary standard for all elements.  Then I analyzed the chlorite standard as a standard:

St  250 Set   1 Chlorite (ideal clinochlore)
TakeOff = 40.0  KiloVolt = 15.0  Beam Current = 30.0  Beam Size =    0
(Magnification (analytical) =   4000),        Beam Mode = Analog  Spot
(Magnification (default) =      200, Magnification (imaging) =    100)
Image Shift (X,Y):                                       -2.00,   3.00

[Mg5Al](Si3Al)O10(OH)8
Number of Data Lines:   5             Number of 'Good' Data Lines:   5
First/Last Date-Time: 11/26/2024 11:20:23 AM to 11/26/2024 11:23:27 AM

Average Total Oxygen:         .000     Average Total Weight%:   99.248
Average Calculated Oxygen:    .000     Average Atomic Number:   10.159
Average Excess Oxygen:        .000     Average Atomic Weight:   15.399
Average ZAF Iteration:        4.00     Average Quant Iterate:     2.00

St  250 Set   1 Chlorite (ideal clinochlore), Results in Elemental Weight Percents
 
ELEM:       Si      Fe      Mg      Al       O       H
TYPE:     ANAL    ANAL    ANAL    ANAL    SPEC    SPEC
BGDS:      LIN     LIN     LIN     LIN
TIME:    10.00   10.00   10.00   10.00     ---     ---
BEAM:    30.00   30.00   30.00   30.00     ---     --- 

ELEM:       Si      Fe      Mg      Al       O       H   SUM  
     6  15.343   -.010  21.286   9.582  51.816   1.451  99.469
     7  15.022    .060  20.422   9.740  51.816   1.451  98.511
     8  15.408    .048  21.492   9.958  51.816   1.451 100.173
     9  15.131    .053  21.802   9.708  51.816   1.451  99.961
    10  15.157    .001  20.113   9.587  51.816   1.451  98.125

AVER:   15.212    .030  21.023   9.715  51.816   1.451  99.248
SDEV:     .159    .032    .722    .153    .000    .000    .897
SERR:     .071    .014    .323    .068    .000    .000
%RSD:     1.05  106.15    3.43    1.58     .00     .00

PUBL:   15.159    n.a.  21.864   9.709  51.816   1.451 100.000
%VAR:      .35     ---   -3.85     .06     .00     .00
DIFF:     .053     ---   -.842    .006    .000    .000
STDS:      160     160     160     160     ---     ---

STKF:    .1621   .0654   .0776   .0335     ---     ---
STCT:    57.10   15.63    1.12    4.62     ---     ---

UNKF:    .1070   .0002   .1511   .0627     ---     ---
UNCT:    37.69     .06    2.17    8.66     ---     ---
UNBG:      .05     .82     .08     .12     ---     ---

ZCOR:   1.4218  1.2253  1.3916  1.5502     ---     ---
KRAW:    .6601   .0038  1.9458  1.8725     ---     ---
PKBG:   826.01    1.07   29.01   73.71     ---     ---

All good.

But note, because this is a standard sample, Probe for EPMA knows to automatically adds in any unanalyzed elements, such as hydrogen!

Next I analyzed the chlorite standard AS AN UNKNOWN sample by making an unknown sample with the SAME NAME as the chlorite standard in this run, because in simulation mode Probe for EPMA will use that standard composition as though it were an actual unknown.

That yields this composition when analyzed after adding oxygen by stoichiometry to the unknown sample:

Un    2 Chlorite (ideal clinochlore)
TakeOff = 40.0  KiloVolt = 15.0  Beam Current = 30.0  Beam Size =    0
(Magnification (analytical) =   4000),        Beam Mode = Analog  Spot
(Magnification (default) =      200, Magnification (imaging) =    100)
Image Shift (X,Y):                                       -2.00,   3.00

[Mg5Al](Si3Al)O10(OH)8

Formula Based on Sum of Cations = .000   Oxygen Calc. by Stoichiometry
Number of Data Lines:   5             Number of 'Good' Data Lines:   5
First/Last Date-Time: 11/26/2024 11:30:20 AM to 11/26/2024 11:33:20 AM

Average Total Oxygen:       39.764     Average Total Weight%:   85.647
Average Calculated Oxygen:  39.764     Average Atomic Number:   10.612
Average Excess Oxygen:        .000     Average Atomic Weight:   20.157
Average ZAF Iteration:        3.00     Average Quant Iterate:     2.00

Oxygen Calculated by Cation Stoichiometry and Included in the Matrix Correction

Un    2 Chlorite (ideal clinochlore), Results in Elemental Weight Percents
 
ELEM:       Si      Fe      Mg      Al       H       O
TYPE:     ANAL    ANAL    ANAL    ANAL    SPEC    CALC
BGDS:      LIN     LIN     LIN     LIN
TIME:    10.00   10.00   10.00   10.00     ---     ---
BEAM:    30.00   30.00   30.00   30.00     ---     ---

ELEM:       Si      Fe      Mg      Al       H       O   SUM  
    11  15.106   -.013  21.420   9.522    .000  39.777  85.811
    12  15.270    .026  20.641   9.612    .000  39.543  85.092
    13  15.294    .014  20.314   9.797    .000  39.516  84.936
    14  15.151   -.009  21.423   9.716    .000  40.005  86.285
    15  15.283    .010  21.042   9.794    .000  39.980  86.109

AVER:   15.221    .005  20.968   9.688    .000  39.764  85.647
SDEV:     .086    .016    .488    .120    .000    .232    .605
SERR:     .039    .007    .218    .054    .000    .104
%RSD:      .57  300.48    2.33    1.24     .00     .58
STDS:      160     160     160     160     ---     ---

STKF:    .1621   .0654   .0776   .0335     ---     ---
STCT:    57.10   15.63    1.12    4.62     ---     ---

UNKF:    .1067   .0000   .1575   .0627     ---     ---
UNCT:    37.57     .01    2.26    8.66     ---     ---
UNBG:      .05     .86     .08     .11     ---     ---

ZCOR:   1.4270  1.1987  1.3314  1.5460     ---     ---
KRAW:    .6580   .0007  2.0284  1.8725     ---     ---
PKBG:   746.56    1.01   29.32   83.57     ---     ---

Un    2 Chlorite (ideal clinochlore), Results in Oxide Weight Percents

ELEM:     SiO2     FeO     MgO   Al2O3     H2O       O   SUM  
    11  32.316   -.017  35.520  17.992    .000    .000  85.811
    12  32.669    .033  34.228  18.161    .000    .000  85.092
    13  32.719    .018  33.688  18.511    .000    .000  84.936
    14  32.414   -.012  35.526  18.357    .000    .000  86.285
    15  32.696    .013  34.894  18.506    .000    .000  86.109

AVER:   32.563    .007  34.771  18.305    .000    .000  85.647
SDEV:     .185    .021    .809    .226    .000    .000    .605
SERR:     .083    .009    .362    .101    .000    .000
%RSD:      .57  300.48    2.33    1.24     .00 -136.93
STDS:      160     160     160     160     ---     ---

What are we still missing?  Well, hydrogen of course. So we need to add hydrogen as an unanalyzed element using the Elements/Cations dialog as seen here:



Note a couple of things, first the hydrogen is an unanalyzed element because the x-ray line is blank. Second, hydrogen (for this material), must be specified as H2O, that is, 2 cations and 1 oxygen.

Next we go into the Calculation Options dialog and specify hydrogen by stoichiometry to stoichiometric oxygen in the proportion that Andrew suggested, that is, 0.4444 H to 1 O:



Now when we click the analyze button we obtain this composition for our "unknown" chlorite:

Un    2 Chlorite (ideal clinochlore)
TakeOff = 40.0  KiloVolt = 15.0  Beam Current = 30.0  Beam Size =    0
(Magnification (analytical) =   4000),        Beam Mode = Analog  Spot
(Magnification (default) =      200, Magnification (imaging) =    100)
Image Shift (X,Y):                                       -2.00,   3.00

[Mg5Al](Si3Al)O10(OH)8

Formula Based on Sum of Cations = .000   Oxygen Calc. by Stoichiometry
Number of Data Lines:   5             Number of 'Good' Data Lines:   5
First/Last Date-Time: 11/26/2024 11:30:20 AM to 11/26/2024 11:33:20 AM

Average Total Oxygen:       51.893     Average Total Weight%:  100.166
Average Calculated Oxygen:  51.893     Average Atomic Number:   10.170
Average Excess Oxygen:        .000     Average Atomic Weight:   15.450
Average ZAF Iteration:        7.00     Average Quant Iterate:     2.00

Oxygen Calculated by Cation Stoichiometry and Included in the Matrix Correction
Element H is Calculated  .444 Atoms Relative To 1.0 Atom of Oxygen

Un    2 Chlorite (ideal clinochlore), Results in Elemental Weight Percents
 
ELEM:       Si      Fe      Mg      Al       H       O
TYPE:     ANAL    ANAL    ANAL    ANAL    STOI    CALC
BGDS:      LIN     LIN     LIN     LIN
TIME:    10.00   10.00   10.00   10.00     ---     ---
BEAM:    30.00   30.00   30.00   30.00     ---     ---

ELEM:       Si      Fe      Mg      Al       H       O   SUM  
    11  15.087   -.014  22.327   9.585   1.447  51.892 100.324
    12  15.256    .027  21.520   9.682   1.451  51.686  99.623
    13  15.280    .014  21.182   9.873   1.453  51.671  99.473
    14  15.132   -.010  22.327   9.781   1.446  52.114 100.789
    15  15.266    .010  21.930   9.864   1.448  52.102 100.621

AVER:   15.204    .006  21.857   9.757   1.449  51.893 100.166
SDEV:     .088    .017    .504    .123    .003    .215    .591
SERR:     .040    .008    .225    .055    .001    .096
%RSD:      .58  300.40    2.31    1.26     .20     .41
STDS:      160     160     160     160     ---     ---

STKF:    .1621   .0654   .0776   .0335     ---     ---
STCT:    57.10   15.63    1.12    4.62     ---     ---

UNKF:    .1067   .0000   .1575   .0627     ---     ---
UNCT:    37.57     .01    2.26    8.66     ---     ---
UNBG:      .05     .86     .08     .11     ---     ---

ZCOR:   1.4255  1.2249  1.3879  1.5569     ---     ---
KRAW:    .6580   .0007  2.0284  1.8725     ---     ---
PKBG:   746.56    1.01   29.32   83.57     ---     ---

Un    2 Chlorite (ideal clinochlore), Results in Oxide Weight Percents

ELEM:     SiO2     FeO     MgO   Al2O3     H2O       O   SUM  
    11  32.276   -.017  37.025  18.110  12.930    .000 100.324
    12  32.638    .034  35.687  18.295  12.969    .000  99.623
    13  32.689    .018  35.127  18.654  12.985    .000  99.473
    14  32.372   -.012  37.024  18.481  12.925    .000 100.789
    15  32.660    .013  36.367  18.637  12.944    .000 100.621

AVER:   32.527    .007  36.246  18.435  12.950    .000 100.166
SDEV:     .189    .022    .836    .233    .026    .000    .591
SERR:     .085    .010    .374    .104    .012    .000
%RSD:      .58  300.40    2.31    1.26     .20  -70.71
STDS:      160     160     160     160     ---     ---

Simulation mode is an excellent way to test these various calculation options in Probe for EPMA. More on using simulation mode to teach EPMA can be found here:

https://smf.probesoftware.com/index.php?topic=837.0

Remember, you can download Probe for EPMA (and CalcZAF) for free to re-process your data from the microprobe or to teach EPMA...

[sckuehn]

Hi Owen,
An EDS TDI (time dependent intensity) correction is a cool idea (why didn't I think of it?), and I don't think the spectrum transfer overhead would be too onerous, assuming the TDI intervals aren't too short.  To me the main difficulty is writing all that code!  It's a lot of work to save each EDS spectrum sub interval and fit each energy channel back to zero time.  It's basically a 1 dimensional version of what I recently got working with the TDI Scanning feature here for x-ray maps:

http://smf.probesoftware.com/index.php?topic=912.0

If we assume 5 or 10 sub intervals, that would mean storing 5 or 10 additional spectra.  The good news is that the TDI correction to the raw spectrum intensities could be applied simply to the energy channels without messing with a background correction.  If I ever get a couple of weeks to work on it, I might look into it.  Maybe after EMAS.
john

The reason my mind went to the "assigned" TDI is that all of that would have to be done only once, rather than at every point, but it sounds like it's the same amount of labor to implement either way. Would the database files get too unwieldy storing 5 spectra per point instead of one?

Zero rush on this, more of a wish than a need. And I agree with your first assertion that having this done by the EDS software is really the best option.

Thanks,
OKN

Yeah, the "assigned" TDI might be slightly less work than just the "self" TDI, but really both "assigned" and "self" TDI methods should be implemented for EDS if we decide to do it. 

So let's say if I can find 3 weeks or so of free time, I will look into it.  It would indeed be cool.  Please remind me in May if you haven't heard back from me by then...
john

I'd love EDS TDI too. Currently, we do lots of beam sensitive volcanic glasses with WDS TDI for Na and K. We do the accompanying Si and Al on EDS with a short 8-10 seconds as the TDI effects are minimal on those elements in such short time. This frees up time on the WDS for other elements, and the EDS intensities are high enough for good precision so fast. EDS TDI would let us do Na, Si, Al by EDS over ~30 sec and improve the precision of all three noticeably.

A simpler but still useful approach would be just having the capacity for two EDS spectra.

That 8 second EDS acquisition we do now wastes the entire rest of a 2 to 6 minute total acquisition depending on the specific method. With a short EDS acquisition followed by a second long one, elements with unstable intensities would go first, and those with stable intensities could go on the the second longer one for much better precision.

Way back around 2009-11, I suggested this kind of approach to some SEM-EDS labs during a round-robin on four glasses that I ran back then. This let them get reasonable sodium concentrations and better precision on minor elements at the same time.

- Steve

[sckuehn]

One more comment on that EDS-TDI:

I'd much rather have PFE handle this than the EDS vendors. Letting the EDS do it makes it too much of a black bock from the PFE side, and there is no way to reprocess differently sometime in the future.

With PFE, switching the type of TDI fit or turning it on and off is something I do a lot to see which is best for a specific sample or batch. That would go away if PFE just received already TDI-corrected results from the EDS.

- Steve

[John Donovan]

We have looked into the idea of acquiring multiple EDS spectra per analysis to perform a TDI correction to the net intensities.

It is a significant amount of work to implement this in code regardless of whether one acquires just two spectra or several or more spectra.

Question: do any EDS vendors provide such a TDI (multiple spectra) EDS acquisition and fitting vs. elapsed time for handling beam sensitive samples, in their own software?