Author Topic: matrix correction comparison  (Read 5569 times)

Ben Buse

  • Professor
  • ****
  • Posts: 414
Re: matrix correction comparison
« Reply #15 on: March 09, 2017, 09:30:38 am »
Hi John and Brian,

Thank you for your replies, that's very interesting - I was curious as two why they did not agree -and you've answered my question. I guess it be possible to modify the MAC30 table with the values from Appendix 5 of green book -but I'm not sure its worth it - plus as Brian says there's difference in fluorescence correction.

As a related question - does the Cameca SX100 PAP routinue use the MAC's as specified by Pouchou and Pichoir - and Reed 1965 fluorescence correction - if so that would explain why the calczaf results never quite match the Cameca software results. I've never been able to find in the Cameca documentation which MAC's they use.

Thanks

Ben

Ben Buse

  • Professor
  • ****
  • Posts: 414
Re: matrix correction comparison
« Reply #16 on: March 09, 2017, 10:04:04 am »
Hi John,

Thanks using your standard.mdb shaw data is working.

The compositions are given in Table 1 of Shaw & Albee 1979? Is that right?

Corresponding as follows

Sample 5 = Kyanite
Sample 6 = wollastonite
Sample 7 = anorthite
Sample 8 = Pyrope
Sample 9 = Grossularite
Sample 10 = Spinel
Sample 11 = Forsterite
Sample 12 = Diopside
Sample 13 = Enstatite
Sample 14 = Enal-20 Glass
Sample 15 = Enal-10 Glass
Sample 16 = Enal-5 Glass
Sample 17 = Glass P-721
Sample 18 = Glass P-722
« Last Edit: March 09, 2017, 10:10:37 am by Ben Buse »

Probeman

  • Emeritus
  • *****
  • Posts: 1955
  • Never sleeps...
    • John Donovan
Re: matrix correction comparison
« Reply #17 on: March 09, 2017, 10:21:54 am »
Thank you for your replies, that's very interesting - I was curious as two why they did not agree -and you've answered my question. I guess it be possible to modify the MAC30 table with the values from Appendix 5 of green book -but I'm not sure its worth it - plus as Brian says there's difference in fluorescence correction.

As a related question - does the Cameca SX100 PAP routinue use the MAC's as specified by Pouchou and Pichoir - and Reed 1965 fluorescence correction - if so that would explain why the calczaf results never quite match the Cameca software results. I've never been able to find in the Cameca documentation which MAC's they use.

Hi Ben,
Yes, as Brian points out, the compositions in the Pouchou database were selected to minimize fluorescence effects.  So the selection of the fluorescence correction shouldn't make too much of a difference. They were mostly interested in the absorption/atomic number corrections (the ZA in the ZAF!). As you know, phi/rho/z methods basically combine the absorption and atomic number effects.

That said, I believe that if you turn off the M-line and beta fluorescence flags in our free CalcZAF utility as found in this dialog:



you will get pretty close to the original Pouchou published results.  It's worth a try...
john
« Last Edit: March 09, 2017, 04:13:13 pm by Probeman »
The only stupid question is the one not asked!

Ben Buse

  • Professor
  • ****
  • Posts: 414
Re: matrix correction comparison
« Reply #18 on: March 13, 2017, 11:20:38 am »
I've run the shaw.dat data - where did the k-ratios come from - did you do the analysis John?.



Also is there an editable version of MAC30 (mac30.dat does not open in text editor), if I get time sometime I might make the PAP amendments - but won't be for awhile.

Ben

Probeman

  • Emeritus
  • *****
  • Posts: 1955
  • Never sleeps...
    • John Donovan
Re: matrix correction comparison
« Reply #19 on: March 13, 2017, 05:12:47 pm »
I've run the shaw.dat data - where did the k-ratios come from - did you do the analysis John?.

Hi Ben,
No I did not.

I believe someone named "Shaw" did these at Cal Tech in the 1970s.  I'm not that old!   ;D    I think John Armstrong or Paul Carpenter would know.

Also is there an editable version of MAC30 (mac30.dat does not open in text editor), if I get time sometime I might make the PAP amendments - but won't be for awhile.

Yes, it's a binary file, but you can edit the MAC30.dat file using CalcZAF.  But even better would be to create your own "user defined" MAC table which you can edit to your heart's content. 

Just use the X-Ray menu in CalcZAF and select the Create Default User Defined MAC Table menu.  It will ask what existing MAC table you want to use as the basis, and you can select MAC30.DAT, and it will then copy it to the name USERMAC.DAT, which you can then edit as much as you want.  This USERMAC.DAT can then be selected in the matrix correction options when you want to test it.
john
« Last Edit: March 13, 2017, 10:05:45 pm by Probeman »
The only stupid question is the one not asked!

Ben Buse

  • Professor
  • ****
  • Posts: 414
Re: matrix correction comparison
« Reply #20 on: March 14, 2017, 12:46:01 am »
Thanks John for the clear explanation of how to do it - I'll give it a go when I get a chance
Ben

Probeman

  • Emeritus
  • *****
  • Posts: 1955
  • Never sleeps...
    • John Donovan
Re: matrix correction comparison
« Reply #21 on: March 14, 2017, 10:54:23 pm »
I've run the shaw.dat data - where did the k-ratios come from - did you do the analysis John?.

Hi Ben,
Paul Carpenter wrote me and mentioned "I don't have many details about the Shaw data, it was collected I am pretty sure on the MAC probe before the Jeol 733 at Caltech. I think maybe it is time for users to generate k-ratio data from the current suite of standards and go from there. This is what I do."

I would agree with Paul here.  I can't remember what you originally wanted the Shaw dataset for, but it is pretty sparse.  This reminds me that I'm hoping some bright young person will start a new k-ratio measurement set to further test our Monte Carlo simulation models, but this time include strongly fluorescing systems also (in addition to strong absorption and atomic number effects).

That was the goal of my "XTREME" (X-ray Table of Ratios with Exceptional Matrix Effects) proposal described here:

http://probesoftware.com/smf/index.php?topic=115.msg426#msg426

john
The only stupid question is the one not asked!

Probeman

  • Emeritus
  • *****
  • Posts: 1955
  • Never sleeps...
    • John Donovan
Re: matrix correction comparison
« Reply #22 on: April 25, 2017, 02:33:48 pm »
Hi John,

Thanks using your standard.mdb shaw data is working.

The compositions are given in Table 1 of Shaw & Albee 1979? Is that right?

Corresponding as follows

Sample 5 = Kyanite
Sample 6 = wollastonite
Sample 7 = anorthite
Sample 8 = Pyrope
Sample 9 = Grossularite
Sample 10 = Spinel
Sample 11 = Forsterite
Sample 12 = Diopside
Sample 13 = Enstatite
Sample 14 = Enal-20 Glass
Sample 15 = Enal-10 Glass
Sample 16 = Enal-5 Glass
Sample 17 = Glass P-721
Sample 18 = Glass P-722

Hi Ben,
So I edited the Shaw.dat file to show the sample names you have listed above (see attached below).   There are quite a few more compositions in the file, so I wonder if Paul Carpenter knows what the rest of these compositions are?
john
The only stupid question is the one not asked!

Ben Buse

  • Professor
  • ****
  • Posts: 414
Re: matrix correction comparison
« Reply #23 on: April 26, 2017, 04:07:52 am »
Hi John,

Sorry I didn't mention - I had the names and ideal compositions from the paper (but the paper does not include the k-ratios - in the shaw.dat file). From the ideal compositions I calculated the difference to the compositions given by the Armstrong and PAP matrix corrections.

The other names are here:

Un1-4. Are the standards MgO, Al2O3, SiO2 and CaO

Un19-Un26 Are glasses P723 to P730

Un27 Cal-Al Pyroxene

And the excel file - of matrix corrections compared is attached. It includes the compositions given in the paper

Ben

Probeman

  • Emeritus
  • *****
  • Posts: 1955
  • Never sleeps...
    • John Donovan
Re: matrix correction comparison
« Reply #24 on: April 26, 2017, 11:27:39 am »
Hi Ben,
Thank-you.

I edited the CalcZAF shaw.dat input file for the proper sample names and I also extracted the "published" compositions from your spreadsheet into a file called shaw.csv, but with the samples listed in the same order as the shaw.dat input file. See attached below (remember everyone, you need to be logged in to see attachments).
john
The only stupid question is the one not asked!

Paul Carpenter

  • Global Moderator
  • Professor
  • *****
  • Posts: 43
    • Washington University Analytical Facilities
Re: matrix correction comparison
« Reply #25 on: April 28, 2017, 08:42:14 am »
I have spent a fair bit of time on this project. The problem with measuring alpha factors is that, for example, in the MgO - Al2O3 binary, there is only one stoichiometric phase MgAl2O4 that defines the midpoint. So there is one measurement only on spinel for Mg and Al relative to MgO and Al2O3 standards. The precision of the measurement has to be very high and when you propagate errors and show the error bars on an alpha factor plot, you realize that there is no justification for using a formulation higher than linear (not constant, not polynomial). Similar problem in other binaries. In MgO - SiO2 there are forsterite and enstatite, but this is the exception. The authors came to this conclusion as well in their search for nonlinearity in the core CMAS system.

Clearly there is a basis for a polynomial formulation when using ZAF prz algorithms where you calculate factors at arbitrarily fine spaced increments of concentration, but that is not the point here.

Secondly. These measurements were made either on an ARL or the MAC probe and are subject to discussion regarding the instrumental stability in the case of the ARL (and takeoff angle not directly comparable to all other measurements made at 40 deg), and in the case of the MAC, non-normal beam incidence. When I was at Caltech I attempted to reconstruct the entire Bence Albee methodology from standards used for measurements and computer output of results. You have to understand that use of these standards predates BSE imaging. The fayalite standard used when I arrived had comments "avoid orange luminescent areas" which were quartz; the fayalite they were using historically was a synthetic material with coprecipitated quartz. The enstatite - Al graduated glasses are of unknown origin and pedigree but are clearly synthetic as well; I could only find two of the three glasses and basically no material left for futher use. Most MgO contains some amount of Ca and so is not pure, strictly speaking, and I doubt that was taken into consideration. There was not a standard mount that had all the standards for the study (so this means that multiple mounts having different carbon coats were used...). This is not to downplay the work that was done, but you have to keep these things in mind.

After the original and only measured alpha factors obtained at 52.5 deg on the ARL were published, they were superseded by the calculated values listed in the Albee Ray paper (and 40 degree factors appeared as well). From there on all factors were calculated by running ZAF in reverse. The desire to analyze S, Cl, etc. motivates a return to a-factors relative to pure element references, i.e., Ziebold Ogilvie, and again does not accomplish much relative to ZAF codes available today.

The real reason for development of BA alpha factors was to eliminate disagreement of ZAF results obtained on the Apollo 11 samples being analyzed by numerous labs, and also to have a small executable program that fit into microcomputers of the day. A BA correction on a Tracor TN2000 took about 1 minute whereas a ZAF correction took about 20 minutes (for an 8 element sample).

Going forward I think we have the same limitations unless one resorts to glasses which could in principle have any intermediate composition in a binary, but are probably goind to exhibit non-binary compositions (ie. a Fe-Si oxide glass could have ferric iron and iron loss to the Pt loop, etc.). The real utility of an alpha factor method is twofold: for graphical comparision of correction magnitude, and for processing X-ray maps (coupled with MAN background method).

I think the method I use is better and that is to compare the measured k-ratio for an element to that calculated by using CalcZAF. A value of kmeas / kcalc = 1 confirms the analytical measurement and all that it depends on (alignment, PHA and deadtime linearity, sample conductivity, etc.) and the correction algorithm (standard composition, algorithm, macs, etc.). Comparison of this measurement as a function of concentration for that element reveals any systematic errors and confirms the internal consistency of the standards used. This goes far beyond the evaluation within a binary join, and also reveals problems with the minor and trace element concentrations being used for microanalysis standards. This unbiased test is really the way to move forward.

Cheers,

Paul
Paul Carpenter
Washington University St. Louis

Probeman

  • Emeritus
  • *****
  • Posts: 1955
  • Never sleeps...
    • John Donovan
Re: matrix correction comparison
« Reply #26 on: April 28, 2017, 11:41:57 am »
Hi Paul,
Thank-you for the historical context and other comments. This is interesting stuff.

To be honest, I'm not sure why Ben was looking at the Shaw data set, but I don't think he was looking at alpha factors as he only mentioned the PAP and Armstrong phi/rho/z methods in his previous posts.  In any event, I figure it's good to get the Shaw dataset into CalcZAF with the proper samples names and published compositions.   If only for purposes of historical documentation!

I agree with your suggestion about using CalcZAF to calculate errors from experimental measurements.  Right now we just have the Pouchou and NIST datasets and they are selected for minimum fluorescence effects and are only binary compounds.  More troubling is that these datasets were measured many years ago (though not as long ago as the Shaw dataset!), so I have to wonder of we should be working on a new experimental dataset of both binary and ternary compositions as I proposed here a while back:

http://probesoftware.com/smf/index.php?topic=115.msg426#msg426

There are a large number of synthetic single crystal materials now commercially available which I suspect we might be able to assume are stoichiometric with fairly high confidence of accuracy, and with these materials perhaps we can start looking more closely at, for example, highly fluorescing systems. And there's still some issues with very large atomic number corrections, e.g., Si Ka in PbSiO3 as just one example.

Karsten and I have looked a little at doing this, but I think a more community based effort might be worth while, even if it were only to obtain some of these new commercially made single crystal materials for round robin measurement of k-ratios under different keV conditions.
john
The only stupid question is the one not asked!

Paul Carpenter

  • Global Moderator
  • Professor
  • *****
  • Posts: 43
    • Washington University Analytical Facilities
Re: matrix correction comparison
« Reply #27 on: April 28, 2017, 04:08:31 pm »
Compositions for En-Al glasses:

St  274 Enstatite 80 Al2O3 20, syn glass, P-405

Syn Schairer, Boyd
Oxide and Elemental Composition

Average Total Oxygen:       47.664     Average Total Weight%:  100.000
Average Calculated Oxygen:  47.664     Average Atomic Number:   10.647
Average Excess Oxygen:        .000     Average Atomic Weight:   20.140

ELEM:      MgO   Al2O3    SiO2       O
XRAY:      ka      ka      ka      ka
OXWT:   32.130  20.000  47.870    .000
ELWT:   19.375  10.585  22.376  47.664
KFAC:    .1441   .0716   .1623   .2540
ZCOR:   1.3443  1.4789  1.3791  1.8769
AT% :   16.055   7.901  16.046  59.998
24 O:    6.422   3.161   6.418  24.000

St  275 Enstatite 90 Al2O3 10, syn glass, P-406

syn Schairer, Boyd
Oxide and Elemental Composition

Average Total Oxygen:       47.738     Average Total Weight%:   99.999
Average Calculated Oxygen:  47.738     Average Atomic Number:   10.647
Average Excess Oxygen:        .000     Average Atomic Weight:   20.109

ELEM:      MgO   Al2O3    SiO2       O
XRAY:      ka      ka      ka      ka
OXWT:   36.139   9.999  53.860    .000
ELWT:   21.793   5.292  25.176  47.738
KFAC:    .1616   .0350   .1864   .2543
ZCOR:   1.3489  1.5126  1.3505  1.8769
AT% :   18.031   3.944  18.026  59.999
24 O:    7.213   1.578   7.210  24.000

St  276 Enstatite 95 Al2O3 5, syn glass, P-407

Syn Schairer, Boyd
Weighed in values (oxide): Mg 38.16 Al 5 Si 56.84
This analysis CIT probe 1997
Oxide and Elemental Composition

Average Total Oxygen:       47.641     Average Total Weight%:   99.724
Average Calculated Oxygen:  47.640     Average Atomic Number:   10.661
Average Excess Oxygen:        .001     Average Atomic Weight:   20.105

ELEM:     Na2O     MgO   Al2O3    SiO2     CaO     FeO       O
XRAY:      ka      ka      ka      ka      ka      ka      ka
OXWT:     .094  37.551   4.930  56.969    .140    .040    .001
ELWT:     .070  22.644   2.609  26.629    .100    .031  47.641
KFAC:    .0004   .1673   .0171   .1996   .0009   .0003   .2529
ZCOR:   1.7065  1.3534  1.5266  1.3343  1.1130  1.2023  1.8834
AT% :     .061  18.783   1.949  19.115    .050    .011  60.030
24 O:     .025   7.509    .779   7.642    .020    .004  24.000


The analysis I made on En95-Al5 again points to small amounts of other elements in some of these standards.

Paul
Paul Carpenter
Washington University St. Louis

Probeman

  • Emeritus
  • *****
  • Posts: 1955
  • Never sleeps...
    • John Donovan
Re: matrix correction comparison
« Reply #28 on: April 29, 2017, 11:44:12 am »
More questions:

Does anyone know how these glasses in the shaw.dat file were characterized for their "published" compositions?   Wet chemistry?

Also, does anyone know if these glasses are still available?  They might be the start of a new ternary (or more) element (XTREME) database with new measurements using new instruments...
john
« Last Edit: May 04, 2017, 01:02:38 pm by John Donovan »
The only stupid question is the one not asked!

Paul Carpenter

  • Global Moderator
  • Professor
  • *****
  • Posts: 43
    • Washington University Analytical Facilities
Re: matrix correction comparison
« Reply #29 on: May 01, 2017, 03:53:43 pm »
The glasses in the Shaw data set with P numbers are almost certainly the Weill (Peggy Dalheim) synthetic glasses in the CMAS compositional system (this means CaO-MgO-Al2O3-SiO2). The P numbers are Caltech standard numbers. The intention was to use them for calorimetric measurements and because accurate composition and homogeneity were important (and a hallmark of Weill's methodology), they are good reference materials. These glasses have been distributed in the past with informal names like "UO Weill glass A" etc.

Thanks for reminding me about the MAC takeoff angle of 38.5 degrees. ARL did not mfg. the MAC probe. I think it stands for Materials Analysis Corporation.

Cheers,

Paul
Paul Carpenter
Washington University St. Louis