Author Topic: Specifying Unanalyzed Elements For a Proper Matrix Correction  (Read 46955 times)

John Donovan

  • Administrator
  • Emeritus
  • *****
  • Posts: 3223
  • Other duties as assigned...
    • Probe Software
Re: Specifying Unanalyzed Elements For a Proper Matrix Correction
« Reply #105 on: August 19, 2019, 04:13:36 PM »
OK, now on to Probe for EPMA.   We think the implementation of the ferrous/ferric calculation (from Droop, 1987) is pretty good, but please download the latest version (12.7.1) and let us know what you think.  The CalcZAF implementation is described here:

https://probesoftware.com/smf/index.php?topic=691.msg8592#msg8592

So let's take an example of hematite. Calculating all iron as FeO we get these results:

Un    3 Hematite, Results in Elemental Weight Percents

ELEM:       Fe       O
TYPE:     ANAL    CALC
BGDS:      LIN
TIME:    10.00     ---
BEAM:    30.00     ---

ELEM:       Fe       O   SUM
    29  68.992  19.766  88.758
    30  69.126  19.805  88.931
    31  68.758  19.699  88.457

AVER:   68.959  19.757  88.715
SDEV:     .186    .053    .240
SERR:     .108    .031
%RSD:      .27     .27
STDS:       39     ---

STKF:    .6810     ---
STCT:   228.06     ---

UNKF:    .6566     ---
UNCT:   219.89     ---
UNBG:      .48     ---

ZCOR:   1.0503     ---
KRAW:    .9642     ---
PKBG:   456.58     ---

Un    3 Hematite, Results in Oxide Weight Percents

ELEM:      FeO       O   SUM
    29  88.758    .000  88.758
    30  88.931    .000  88.931
    31  88.457    .000  88.457

AVER:   88.715    .000  88.715
SDEV:     .240    .000    .240
SERR:     .138    .000
%RSD:      .27  -91.65
STDS:       39     ---

Note the low totals (~88%) due to the missing ferric oxygen.  Now we specify the ferrous/ferric calculation and the mineral cations and oxygen (2 and 3 respectively for hematites) as seen here:



and re-calculating we now obtain these results:

Un    3 Hematite, Results in Elemental Weight Percents
 
ELEM:       Fe       O
TYPE:     ANAL    CALC
BGDS:      LIN
TIME:    10.00     ---
BEAM:    30.00     ---

ELEM:       Fe       O   SUM
    29  70.168  30.155 100.323
    30  70.305  30.213 100.518
    31  69.930  30.052  99.982

AVER:   70.134  30.140 100.274
SDEV:     .190    .081    .271
SERR:     .109    .047
%RSD:      .27     .27
STDS:       39     ---

STKF:    .6810     ---
STCT:   228.06     ---

UNKF:    .6566     ---
UNCT:   219.89     ---
UNBG:      .48     ---

ZCOR:   1.0682     ---
KRAW:    .9642     ---
PKBG:   456.58     ---

Ferrous/Ferric Calculation Results:
        Ferric/TotalFe   FeO        Fe2O3    Oxygen from Fe2O3
    29       1.000        .000     100.323      10.052
    30       1.000        .000     100.518      10.071
    31       1.000        .000      99.982      10.017

AVER:        1.000        .000     100.274      10.047

Un    3 Hematite, Results in Oxide Weight Percents

ELEM:      FeO       O   SUM
    29  90.271  10.052 100.323
    30  90.447  10.071 100.518
    31  89.965  10.017  99.982

AVER:   90.228  10.047 100.274
SDEV:     .244    .027    .271
SERR:     .141    .016
%RSD:      .27     .27
STDS:       39     ---

Our totals are now much better but because we included this excess oxygen in the matrix correction as suggested by Brian Joy, even more impressive is that our Fe went from 68.959 to 70.134 wt%!

We also added this ferrous/ferric output to the User Specified Output options as seen here:



so now these results can be output to Excel as usual:

« Last Edit: August 19, 2019, 05:56:38 PM by John Donovan »
John J. Donovan, Pres. 
(541) 343-3400

"Not Absolutely Certain, Yet Reliable"

Probeman

  • Emeritus
  • *****
  • Posts: 2771
  • Never sleeps...
    • John Donovan
Re: Specifying Unanalyzed Elements For a Proper Matrix Correction
« Reply #106 on: September 11, 2019, 02:38:22 PM »
Here's a "real world" example of a titanium magnetite with the ferrous/ferric excess oxygen calculation in Probe for EPMA using the method of Droop (1987) to calculate the excess oxygen from ferric iron. This excess oxygen is then added into the matrix correction for improved accuracy. The output below also includes oxides and sum of cations also as specified in the Calculation Options dialog.

Un   30 8400 Mgt Trav In-Out
TakeOff = 40.0  KiloVolt = 15.0  Beam Current = 50.0  Beam Size =    0
(Magnification (analytical) =  20000),        Beam Mode = Analog  Spot
(Magnification (default) =     1000, Magnification (imaging) =    100)
Image Shift (X,Y):                                       -2.00,   3.00

Formula Based on Sum of Cations = 3.00   Oxygen Calc. by Stoichiometry
Number of Data Lines:   6             Number of 'Good' Data Lines:   6
First/Last Date-Time: 07/23/2019 02:12:20 PM to 07/23/2019 02:29:56 PM

Average Total Oxygen:       28.684     Average Total Weight%:  100.294
Average Calculated Oxygen:  28.684     Average Atomic Number:   20.135
Average Excess Oxygen:        .000     Average Atomic Weight:   31.696
Average ZAF Iteration:        3.00     Average Quant Iterate:     4.00

Oxygen Calculated by Cation Stoichiometry and Included in the Matrix Correction
Excess Oxygen From Ferric Iron Calculated and Included in the Matrix Correction
Charge Balance Method of Droop (1987), Total Cations= 3.00, Total Oxygens= 4.00

Un   30 8400 Mgt Trav In-Out, Results in Elemental Weight Percents
 
ELEM:       Fe      Mg      Si      Ti       V      Mn      Cr      Al       O
TYPE:     ANAL    ANAL    ANAL    ANAL    ANAL    ANAL    ANAL    ANAL    CALC
BGDS:      MAN     MAN     MAN     MAN     LIN     LIN     MAN     MAN
TIME:    40.00  140.00   40.00   80.00   30.00   30.00   88.00  141.00     ---
BEAM:    50.32   50.32   50.32   50.32   50.32   50.32   50.32   50.32     ---

ELEM:       Fe      Mg      Si      Ti       V      Mn      Cr      Al       O   SUM 
    72  60.751   1.862    .019   6.903    .309    .403    .054   1.270  28.664 100.234
    73  60.644   1.863    .075   6.872    .333    .407    .059   1.258  28.657 100.170
    74  60.946   1.871    .021   6.884    .295    .405    .056   1.255  28.732 100.464
    75  60.505   1.856    .028   6.918    .323    .420    .056   1.260  28.584  99.950
    76  60.954   1.876    .016   6.951    .335    .422    .043   1.251  28.776 100.623
    77  60.805   1.858    .023   6.949    .301    .404    .044   1.249  28.693 100.325

AVER:   60.768   1.864    .030   6.913    .316    .410    .052   1.257  28.684 100.294
SDEV:     .175    .008    .022    .032    .017    .008    .007    .008    .066    .235
SERR:     .071    .003    .009    .013    .007    .003    .003    .003    .027
%RSD:      .29     .41   73.90     .47    5.40    2.06   12.65     .61     .23
STDS:      395      12      14      22      23      25      24     396     ---

STKF:    .6779   .4736   .4101   .5547   .6328   .7341   .6400   .0469     ---
STCT:   6812.2 53482.2  7089.5 14540.1 13304.5 22081.3  4697.4  3264.8     ---

UNKF:    .5649   .0085   .0002   .0703   .0033   .0038   .0006   .0072     ---
UNCT:   5676.3   964.8     3.7  1841.7    69.6   114.7     4.3   503.0     ---
UNBG:     20.5    79.1     5.5    52.0    25.2    49.5     9.6    48.0     ---

ZCOR:   1.0758  2.1821  1.4034   .9839   .9547  1.0761   .8885  1.7401     ---
KRAW:    .8333   .0180   .0005   .1267   .0052   .0052   .0009   .1541     ---
PKBG:   277.35   13.19    1.68   36.42    3.77    3.32    1.45   11.48     ---
INT%:      .00    ----    ----    ----  -17.39    -.03  -10.90     .00     ---

Ferrous/Ferric Calculation Results:
        Ferric/TotalFe   FeO        Fe2O3    Oxygen from Fe2O3 
    72        .458      42.390      39.749       3.983
    73        .455      42.482      39.494       3.957
    74        .460      42.324      40.100       4.018
    75        .456      42.333      39.459       3.954
    76        .458      42.468      39.953       4.003
    77        .458      42.388      39.828       3.990

AVER:         .458      42.397      39.764       3.984

Un   30 8400 Mgt Trav In-Out, Results in Oxide Weight Percents

ELEM:      FeO     MgO    SiO2    TiO2    V2O3     MnO   Cr2O3   Al2O3       O   SUM 
    72  78.156   3.087    .041  11.515    .454    .520    .079   2.400   3.983 100.234
    73  78.019   3.090    .161  11.464    .491    .526    .086   2.377   3.957 100.170
    74  78.407   3.103    .044  11.484    .433    .523    .081   2.371   4.018 100.464
    75  77.839   3.078    .059  11.539    .476    .542    .082   2.382   3.954  99.950
    76  78.418   3.110    .034  11.594    .492    .545    .063   2.363   4.003 100.623
    77  78.225   3.080    .050  11.591    .442    .522    .065   2.359   3.990 100.325

AVER:   78.177   3.091    .065  11.531    .465    .530    .076   2.375   3.984 100.294
SDEV:     .225    .013    .048    .054    .025    .011    .010    .015    .025    .235
SERR:     .092    .005    .020    .022    .010    .004    .004    .006    .010
%RSD:      .29     .41   73.90     .47    5.40    2.06   12.65     .61     .64
STDS:      395      12      14      22      23      25      24     396     ---

Un   30 8400 Mgt Trav In-Out, Results Based on Sum of 3 Cations

ELEM:       Fe      Mg      Si      Ti       V      Mn      Cr      Al       O   SUM 
    72   2.381    .168    .001    .315    .013    .016    .002    .103   3.921   6.921
    73   2.377    .168    .006    .314    .014    .016    .002    .102   3.921   6.921
    74   2.384    .168    .002    .314    .013    .016    .002    .102   3.922   6.922
    75   2.378    .168    .002    .317    .014    .017    .002    .103   3.921   6.921
    76   2.380    .168    .001    .316    .014    .017    .002    .101   3.922   6.922
    77   2.382    .167    .002    .317    .013    .016    .002    .101   3.923   6.923

AVER:    2.380    .168    .002    .316    .014    .016    .002    .102   3.922   6.922
SDEV:     .002    .000    .002    .001    .001    .000    .000    .001    .001    .001
SERR:     .001    .000    .001    .001    .000    .000    .000    .000    .000
%RSD:      .10     .24   73.96     .46    5.41    2.06   12.76     .74     .02

Note that if we did *not* include the calculation of excess oxygen in the matrix correction our oxides and formula would look like this:

Un   30 8400 Mgt Trav In-Out, Results in Oxide Weight Percents

ELEM:      FeO     MgO    SiO2    TiO2    V2O3     MnO   Cr2O3   Al2O3       O   SUM 
    72  77.687   3.107    .037  11.434    .451    .517    .074   2.412    .000  95.718
    73  77.554   3.109    .158  11.383    .487    .523    .080   2.388    .000  95.684
    74  77.933   3.123    .041  11.402    .430    .520    .076   2.383    .000  95.908
    75  77.374   3.098    .056  11.458    .472    .539    .076   2.393    .000  95.467
    76  77.946   3.130    .031  11.512    .489    .542    .058   2.375    .000  96.083
    77  77.755   3.100    .047  11.509    .439    .519    .059   2.371    .000  95.799

AVER:   77.708   3.111    .062  11.450    .461    .527    .071   2.387    .000  95.776
SDEV:     .221    .013    .048    .054    .025    .011    .010    .015    .000    .210
SERR:     .090    .005    .020    .022    .010    .004    .004    .006    .000
%RSD:      .28     .41   78.00     .47    5.41    2.06   13.52     .61 -167.33
STDS:      395      12      14      22      23      25      24     396     ---

Un   30 8400 Mgt Trav In-Out, Results Based on Sum of 3 Cations

ELEM:       Fe      Mg      Si      Ti       V      Mn      Cr      Al       O   SUM 
    72   2.379    .170    .001    .315    .013    .016    .002    .104   3.376   6.376
    73   2.375    .170    .006    .313    .014    .016    .002    .103   3.379   6.379
    74   2.382    .170    .002    .313    .013    .016    .002    .103   3.374   6.374
    75   2.376    .170    .002    .316    .014    .017    .002    .104   3.378   6.378
    76   2.378    .170    .001    .316    .014    .017    .002    .102   3.376   6.376
    77   2.380    .169    .002    .317    .013    .016    .002    .102   3.377   6.377

AVER:    2.378    .170    .002    .315    .014    .016    .002    .103   3.377   6.377
SDEV:     .002    .000    .002    .001    .001    .000    .000    .001    .002    .002
SERR:     .001    .000    .001    .001    .000    .000    .000    .000    .001
%RSD:      .10     .24   78.05     .46    5.41    2.06   13.63     .74     .06

Of course the oxygen formula is low, but note the effect of this "missing oxygen" on the concentrations of the other elements. E.g., without the excess oxygen in the matrix correction, the Fe and Ti concentrations went down, but the Mg and Al concentrations went up!
« Last Edit: September 11, 2019, 03:26:57 PM by Probeman »
The only stupid question is the one not asked!

Probeman

  • Emeritus
  • *****
  • Posts: 2771
  • Never sleeps...
    • John Donovan
Re: Specifying Unanalyzed Elements For a Proper Matrix Correction
« Reply #107 on: February 10, 2020, 10:14:23 PM »
Mike Dungan, Andrew Locock and I fixed a problem with some oxide compositions in the ferric/ferric excess oxygen calculation as seen here:

Ferrous/Ferric Calculation Results:
        Ferric/TotalFe   FeO        Fe2O3    Oxygen from Fe2O3 
    37        .387      44.824      31.498       3.156
    38        .387      44.633      31.339       3.140

AVER:         .387      44.729      31.419       3.148

Un    6 CH-19_Ox99_Mgt Core-1, Results in Oxide Weight Percents

ELEM:      FeO     MgO    TiO2    V2O3     MnO   Cr2O3   Al2O3    SiO2     CaO       O   SUM 
    37  73.166   2.152  18.583    .543    .586    .006   1.835    .053    .014   3.156 100.093
    38  72.832   2.159  18.500    .553    .574    .002   1.906    .043    .014   3.140  99.723

AVER:   72.999   2.156  18.541    .548    .580    .004   1.870    .048    .014   3.148  99.908
SDEV:     .236    .005    .059    .008    .008    .003    .051    .007    .000    .011    .261
SERR:     .167    .003    .042    .005    .006    .002    .036    .005    .000    .008
%RSD:      .32     .23     .32    1.39    1.45   69.25    2.70   14.77    1.72     .36
STDS:      395      12      22      23      25      24     306      14     306     ---

Un    6 CH-19_Ox99_Mgt Core-1, Results Based on Sum of 3 Cations

ELEM:       Fe      Mg      Ti       V      Mn      Cr      Al      Si      Ca       O   SUM 
    37   2.251    .118    .514    .016    .018    .000    .080    .002    .001   4.000   7.000
    38   2.248    .119    .514    .016    .018    .000    .083    .002    .001   4.000   7.000

AVER:    2.250    .118    .514    .016    .018    .000    .081    .002    .001   4.000   7.000
SDEV:     .002    .001    .000    .000    .000    .000    .002    .000    .000    .000    .000
SERR:     .002    .000    .000    .000    .000    .000    .002    .000    .000    .000
%RSD:      .10     .46     .09    1.62    1.22   69.07    2.93   14.55    1.49     .00

This change now properly handles oxides such as Al2O3, V2O3, Cr2O3, etc. You can update PFE anytime.
« Last Edit: February 11, 2020, 07:32:25 AM by Probeman »
The only stupid question is the one not asked!

John Donovan

  • Administrator
  • Emeritus
  • *****
  • Posts: 3223
  • Other duties as assigned...
    • Probe Software
Re: Specifying Unanalyzed Elements For a Proper Matrix Correction
« Reply #108 on: June 29, 2020, 09:34:18 AM »
We've recently been working with Julien Allaz to fix the way in which the oxygen-halogen correction is applied to standards in Probe for EPMA as first reported by Ben Wade:

https://probesoftware.com/smf/index.php?topic=1247.msg9305#msg9305

During that time we realized that these recent halogen correction issues for standards (when oxygen is calculated by stoichiometry rather than a fixed concentration), are somewhat similar to the excess oxygen calculation using the Droop method of charge balance (hey- it's all about charge balance!). Again only for standard samples, when oxygen is calculated by stoichiometry (as opposed to added by fixed concentration from the standard database).

Previously, for the excess oxygen from ferric iron calculation, we simply did not allow one to perform this excess oxygen calculation for standards (only for unknown samples). But we realized yesterday that it's the same issue in that a double correction is applied for standards if the excess (or deficit) oxygen is already included in the standard database, and the excess (or deficit) oxygen correction for ferric oxygen (or halogen equivalence) is applied in Probe for EPMA, *and* the oxygen in the standard is calculated by stocihiometry.  Whew!

So what we did was enable the calculation of excess oxygen for standards (just as we have had for the halogen correction), so if the standard analysis is using the default fixed oxygen concentration from the standard database, all is well.  This can be seen is this output:

Excess Oxygen From Ferric Iron was not Included in the Matrix Correction (because oxygen was not calculated by cation stoichiometry)

St  395 Set   1 Magnetite U.C. #3380, Results in Elemental Weight Percents
 
ELEM:        F      Fe      Al      Mg      Mn       O
TYPE:     ANAL    ANAL    SPEC    SPEC    SPEC    SPEC
BGDS:      LIN     LIN
TIME:    20.00   20.00     ---     ---     ---     ---
BEAM:    29.98   29.98     ---     ---     ---     ---

ELEM:        F      Fe      Al      Mg      Mn       O   SUM 
    10    .051  71.844    .201    .072    .054  27.803 100.025
    11    .073  72.195    .201    .072    .054  27.803 100.398
    12    .026  72.209    .201    .072    .054  27.803 100.365

AVER:     .050  72.083    .201    .072    .054  27.803 100.263
SDEV:     .023    .207    .000    .000    .000    .000    .207
SERR:     .014    .119    .000    .000    .000    .000
%RSD:    46.70     .29     .00     .00     .00     .00

PUBL:     n.a.  72.080    .201    .072    .054  27.803 100.210
%VAR:      ---   (.00)     .00     .00     .00     .00
DIFF:      ---   (.00)    .000    .000    .000    .000
STDS:      835     395     ---     ---     ---     ---

STKF:    .1715   .6779     ---     ---     ---     ---
STCT:    57.18  228.22     ---     ---     ---     ---

UNKF:    .0002   .6779     ---     ---     ---     ---
UNCT:      .07  228.22     ---     ---     ---     ---
UNBG:      .32     .49     ---     ---     ---     ---

ZCOR:   2.2854  1.0633     ---     ---     ---     ---
KRAW:    .0013  1.0000     ---     ---     ---     ---
PKBG:     1.23  466.16     ---     ---     ---     ---

Ferrous/Ferric Calculation Results:
        Ferric/TotalFe   FeO        Fe2O3    Oxygen from Fe2O3 
    10        .000        .000        .000        .000
    11        .000        .000        .000        .000
    12        .000        .000        .000        .000

AVER:         .000        .000        .000        .000

But in the case where the user changes the standard analysis calculation to calculate oxygen by stoichiometry, the program will now zero out the excess oxygen (just as it now does for the deficit oxygen from halogens), and not perform a double correction as seen here:

Oxygen Calculated by Cation Stoichiometry and Included in the Matrix Correction
Oxygen Equivalent from Halogens (F/Cl/Br/I) was Subtracted in the Matrix Correction
Excess Oxygen From Ferric Iron Calculated and Included in the Matrix Correction
Charge Balance Method of Droop (1987), Total Cations= 3.00, Total Oxygens= 4.00

St  395 Set   1 Magnetite U.C. #3380, Results in Elemental Weight Percents
 
ELEM:        F      Fe      Al      Mg      Mn       O
TYPE:     ANAL    ANAL    SPEC    SPEC    SPEC    CALC
BGDS:      LIN     LIN
TIME:    20.00   20.00     ---     ---     ---     ---
BEAM:    29.98   29.98     ---     ---     ---     ---

ELEM:        F      Fe      Al      Mg      Mn       O   SUM 
    10    .051  71.838    .201    .072    .054  27.742  99.957
    11    .073  72.206    .201    .072    .054  27.908 100.514
    12    .026  72.215    .201    .072    .054  27.859 100.427

AVER:     .050  72.086    .201    .072    .054  27.836 100.299
SDEV:     .023    .215    .000    .000    .000    .085    .300
SERR:     .014    .124    .000    .000    .000    .049
%RSD:    46.75     .30     .00     .00     .00     .31

PUBL:     n.a.  72.080    .201    .072    .054  27.803 100.210
%VAR:      ---   (.01)     .00     .00     .00     .12
DIFF:      ---   (.01)    .000    .000    .000    .033
STDS:      835     395     ---     ---     ---     ---

STKF:    .1715   .6779     ---     ---     ---     ---
STCT:    57.18  228.22     ---     ---     ---     ---

UNKF:    .0002   .6779     ---     ---     ---     ---
UNCT:      .07  228.22     ---     ---     ---     ---
UNBG:      .32     .49     ---     ---     ---     ---

ZCOR:   2.2865  1.0633     ---     ---     ---     ---
KRAW:    .0013  1.0000     ---     ---     ---     ---
PKBG:     1.23  466.16     ---     ---     ---     ---

Ferrous/Ferric Calculation Results:
        Ferric/TotalFe   FeO        Fe2O3    Oxygen from Fe2O3 
    10        .674      30.092      69.266       6.940
    11        .678      29.937      69.966       7.010
    12        .671      30.591      69.252       6.938

AVER:         .674      30.207      69.495       6.963


So now, the code for dealing with this is seen here:

Code: [Select]
If sample(1).Type% = 1 Then
analysis.WtPercents!(chan%) = ConvertTotalToExcessOxygen!(Int(1), sample(), stdsample())
If UseOxygenFromHalogensCorrectionFlag And sample(1).OxideOrElemental% = 1 Then
If analysis.WtPercents!(chan%) < 0# Then analysis.WtPercents!(chan%) = 0#  ' zero out oxygen deficit from standard database
End If
If sample(1).FerrousFerricCalculationFlag And sample(1).OxideOrElemental% = 1 Then
If analysis.WtPercents!(chan%) > 0# Then analysis.WtPercents!(chan%) = 0#  ' zero out oxygen excess from standard database
End If

' For unknowns, use specified oxygen weight percent
Else
analysis.WtPercents!(chan%) = stdsample(1).ElmPercents!(ip%)
End If

One possible remaining issue is when one has both halogens and ferric iron present in a standard, and one tries to calculate oxygen by stoichiometry, the excess/deficit oxygen might not be handled perfectly. In this case, we would just say, simply use the default to calculate oxygen elementally (fixed concentration from the standard database) and all will be well.  But I think it will work, because if both the oxygen-halogen correction *and* the ferric iron calculation correction are turned on, you'll get a zero value for excess/deficit oxygen, so it should all work.

Remember, in all cases, the calculations for unknown samples are handled just fine. This is all pretty complicated we know, but please ask us any questions you may have.

In the meantime, update Probe for EPMA from the Help menu and these options are all available now.
« Last Edit: June 29, 2020, 10:08:29 PM by John Donovan »
John J. Donovan, Pres. 
(541) 343-3400

"Not Absolutely Certain, Yet Reliable"

John Donovan

  • Administrator
  • Emeritus
  • *****
  • Posts: 3223
  • Other duties as assigned...
    • Probe Software
Re: Specifying Unanalyzed Elements For a Proper Matrix Correction
« Reply #109 on: August 27, 2021, 08:19:43 AM »
Recently Andrew Locock found a mineral calculation for an unanalyzed elements situation which did not properly apply the oxygen-halogen correction when calculating oxygen by stoichiometry (that is, when not measuring oxygen, or specifying oxygen as a fixed concentration).  The following examples were all measured by Andrew on one of his microprobe instruments (yes he has two of them!).

I should also mention that a full paper on the effects of unanalyzed elements in EPMA analysis is being prepared with Aurelien Moy as first author (and Andrew and myself and several others) that will be submitted later this year for publication.

As a quick review of the unanalyzed element problem can be seen is this carbonate example where oxygen is calculated by stoichiometry and carbon is calculated relative to stoichiometric oxygen in the ratio of 0.333 atoms of carbon to one atom of oxygen:



One can click on the images to see them better, though I'm not quite sure why he specified 0.33368 atoms of carbon in this example...

Another slightly more complicated example is that of tourmaline where 0.129 atoms of hydrogen are calculated relative to stoichiometric oxygen *and* 0.5 atoms of boron are calculated relative to silicon:



Now as for the halogen-oxygen correction, this is applied when oxygen is calculated by stoichiometry and a halogen element (chlorine, fluorine, etc.) is also present and replacing some of the stoichiometric oxygen. In these cases, unless a correction is applied to the stoichiometric oxygen concentration, too much oxygen will be included in the EPMA matrix iteration and the correction of the other elements will be calculated incorrectly.

For chlorine measurements the effect of this excess oxygen is fairly small, but for fluorine, which is heavily absorbed by oxygen, the effect is much larger. There is a separate topic devoted to this topic here, if anyone wants to learn more about these details:

https://probesoftware.com/smf/index.php?topic=1247.0

So back to Andrew's observations. What Andrew found is that when he was *not* measuring fluorine and was calculating oxygen by stoichiometry in some topaz minerals, the halogen-oxygen correction was being applied correctly as seen here:



That is, when the unanalyzed fluorine is calculated by stoichiometry to stoichiometric oxygen.

However, he found that when he tried to calculate the fluorine by stoichiometry to aluminum (aluminium to some!), the program did not apply the halogen-oxygen correction as expected. This was because it turned out that the element relative to another element code was located *after* the oxygen by stoichiometry code in the matrix iteration.  When we moved the element relative to another element code to *before* the oxygen by stoichiometry code, it all worked as seen here:



 8)

Any questions? 
John J. Donovan, Pres. 
(541) 343-3400

"Not Absolutely Certain, Yet Reliable"

Probeman

  • Emeritus
  • *****
  • Posts: 2771
  • Never sleeps...
    • John Donovan
Re: Specifying Unanalyzed Elements For a Proper Matrix Correction
« Reply #110 on: August 27, 2021, 09:55:31 AM »
I have a question: is it an "oxygen-halogen" correction, or a "halogen-oxygen" correction?   :D
The only stupid question is the one not asked!

AndrewLocock

  • Professor
  • ****
  • Posts: 98
    • University of Alberta Electron Microprobe Laboratory
Re: Specifying Unanalyzed Elements For a Proper Matrix Correction
« Reply #111 on: August 27, 2021, 10:21:42 AM »
I have a question: is it an "oxygen-halogen" correction, or a "halogen-oxygen" correction?   :D

"Oxygen equivalent of fluorine" is how it was put by Deer, Howie and Zussman in the Appendix of their book
An Introduction to the Rock-Forming Minerals.

So, more broadly, the oxygen equivalent of the halogen content.

Sorry about the typo in the carbonate example above.
Andrew

John Donovan

  • Administrator
  • Emeritus
  • *****
  • Posts: 3223
  • Other duties as assigned...
    • Probe Software
Re: Specifying Unanalyzed Elements For a Proper Matrix Correction
« Reply #112 on: August 27, 2021, 12:02:57 PM »
I have a question: is it an "oxygen-halogen" correction, or a "halogen-oxygen" correction?   :D

"Oxygen equivalent of fluorine" is how it was put by Deer, Howie and Zussman in the Appendix of their book
An Introduction to the Rock-Forming Minerals.

So, more broadly, the oxygen equivalent of the halogen content.

Sorry about the typo in the carbonate example above.
Andrew

OK, then I'm going to use "oxygen-halogen" correction, because I can't usually spell "equivalent" or "equivalence"!    ;D
« Last Edit: August 27, 2021, 12:44:20 PM by John Donovan »
John J. Donovan, Pres. 
(541) 343-3400

"Not Absolutely Certain, Yet Reliable"

Probeman

  • Emeritus
  • *****
  • Posts: 2771
  • Never sleeps...
    • John Donovan
Re: Specifying Unanalyzed Elements For a Proper Matrix Correction
« Reply #113 on: August 29, 2021, 09:32:15 AM »
I think it's worth demonstrating in some detail exactly how important it is to include these unanalyzed elements in the matrix correction. For example here are the quant results for the above topaz example (data from Locock) with fluorine not included in the matrix iteration:

Un    2 Topaz, Results in Elemental Weight Percents
 
ELEM:       Al      Si       F       O
TYPE:     ANAL    ANAL    SPEC    CALC
BGDS:      LIN     LIN
TIME:    20.00   20.00     ---     ---
BEAM:    15.18   15.18     ---     ---

ELEM:       Al      Si       F       O   SUM 
     6  28.266  15.337    .000  42.617  86.220
     7  28.270  15.327    .000  42.608  86.205
     8  28.137  15.383    .000  42.555  86.075
     9  28.351  15.455    .000  42.827  86.634
    10  28.221  15.368    .000  42.611  86.200

AVER:   28.249  15.374    .000  42.644  86.267
SDEV:     .078    .051    .000    .106    .213
SERR:     .035    .023    .000    .047
%RSD:      .28     .33     .00     .25
STDS:       73      73     ---     ---

STKF:    .2305   .1076     ---     ---
STCT:   448.80  238.86     ---     ---

UNKF:    .2303   .1078     ---     ---
UNCT:   448.33  239.45     ---     ---
UNBG:     1.83    1.54     ---     ---

ZCOR:   1.2268  1.4258     ---     ---
KRAW:    .9990  1.0025     ---     ---
PKBG:   245.93  156.66     ---     ---

And here is with the fluorine included:

Un    2 Topaz, Results in Elemental Weight Percents
 
ELEM:       Al      Si       F       O
TYPE:     ANAL    ANAL    STOI    CALC
BGDS:      LIN     LIN
TIME:    20.00   20.00     ---     ---
BEAM:    15.18   15.18     ---     ---

ELEM:       Al      Si       F       O   SUM 
     6  29.308  15.258  20.647  34.758  99.970
     7  29.312  15.247  20.645  34.750  99.954
     8  29.176  15.304  20.641  34.696  99.817
     9  29.393  15.376  20.675  34.957 100.401
    10  29.262  15.288  20.647  34.752  99.948

AVER:   29.290  15.294  20.651  34.783 100.018
SDEV:     .080    .051    .014    .101    .223
SERR:     .036    .023    .006    .045
%RSD:      .27     .33     .07     .29
STDS:       73      73     ---     ---

STKF:    .2305   .1076     ---     ---
STCT:   448.80  238.86     ---     ---

UNKF:    .2303   .1078     ---     ---
UNCT:   448.33  239.45     ---     ---
UNBG:     1.83    1.54     ---     ---

ZCOR:   1.2720  1.4184     ---     ---
KRAW:    .9990  1.0025     ---     ---
PKBG:   245.93  156.66     ---     ---

Note that the Al concentration changes by around 1 wt% *absolute* or over 3% relative!

And if we don't specify the oxygen-halogen correction to remove the excess stoichiometric oxygen from the matrix iteration, not only is our total way too high but look at the effect on the Al concentration from including that excess oxygen:

Un    2 Topaz, Results in Elemental Weight Percents
 
ELEM:       Al      Si       F       O
TYPE:     ANAL    ANAL    STOI    CALC
BGDS:      LIN     LIN
TIME:    20.00   20.00     ---     ---
BEAM:    15.18   15.18     ---     ---

ELEM:       Al      Si       F       O   SUM 
     6  29.819  15.136  23.211  43.768 111.934
     7  29.823  15.125  23.210  43.760 111.918
     8  29.685  15.182  23.209  43.702 111.778
     9  29.904  15.254  23.233  43.979 112.369
    10  29.772  15.166  23.212  43.761 111.911

AVER:   29.801  15.173  23.215  43.794 111.982
SDEV:     .080    .051    .010    .106    .225
SERR:     .036    .023    .004    .048
%RSD:      .27     .34     .04     .24
STDS:       73      73     ---     ---

STKF:    .2305   .1076     ---     ---
STCT:   448.80  238.86     ---     ---

UNKF:    .2303   .1078     ---     ---
UNCT:   448.33  239.45     ---     ---
UNBG:     1.83    1.54     ---     ---

ZCOR:   1.2942  1.4071     ---     ---
KRAW:    .9990  1.0025     ---     ---
PKBG:   245.93  156.66     ---     ---

Bottom line: whether the element is missing or in excess, it needs to be dealt with the the matrix iteration for accurate results!
« Last Edit: November 05, 2021, 01:29:17 PM by Probeman »
The only stupid question is the one not asked!

Probeman

  • Emeritus
  • *****
  • Posts: 2771
  • Never sleeps...
    • John Donovan
Re: Specifying Unanalyzed Elements For a Proper Matrix Correction
« Reply #114 on: November 05, 2021, 04:52:24 PM »
In this topic we have been emphasizing the importance of including the concentrations of all unanalyzed (specified) elements to obtain an accurate matrix correction, and therefore hopefully, accurate results:

https://www.cambridge.org/core/journals/microscopy-and-microanalysis/article/epma-matrix-correction-all-elements-must-be-present-for-accuracy-four-examples-with-b-c-o-and-f/759C69631A916EA69913539C22989A31

This is because for major elements the most important parameters for analysis accuracy are the accuracy of the matrix correction *and* the accuracy of the standard compositions. For major elements, the accuracy of the background measurement is usually "in the noise", and usually, little affected by spectral interferences.

But what about trace elements? For trace element accuracy, the accuracy of the matrix corrections and standard compositions are not so critical, as it is the accuracy of the background measurements (and spectral interference correction!) that becomes the dominant factor:

https://probesoftware.com/smf/index.php?topic=928.msg8498#msg8498

We can appreciate this point by assuming typical EPMA matrix correction accuracy of say ~2%. If we are measuring a concentration of say, 1000 PPM (0.1 wt%), then 2% of 1000 PPM is 20 PPM (0.002 wt%). A level of accuracy very likely close to the precision of our measurement. Again, usually "in the noise" of our measurement.

We can demonstrate this more clearly in a different manner by making trace element measurements in a material of known composition, the major (and minor) elements being specified to account for the effect of the matrix effects of the unanalyzed elements. Here is the analysis with *no* (unanalyzed) matrix specified:

Un   33 MA-1058 Rxn-1 (trav), Results in Elemental Weight Percents
 
ELEM:       Zr      Nb      La      Sr      Ti      Si      Ca      Al      Fe      Na       K      Mg      Mn      Cr       O
TYPE:     ANAL    ANAL    ANAL    ANAL    ANAL    SPEC    SPEC    SPEC    SPEC    SPEC    SPEC    SPEC    SPEC    SPEC    SPEC
BGDS:     MULT    MULT    MULT    MULT    MULT
TIME:   240.00  240.00  240.00  240.00  240.00     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---
BEAM:    80.06   80.06   80.06   80.06   80.06     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---

ELEM:       Zr      Nb      La      Sr      Ti      Si      Ca      Al      Fe      Na       K      Mg      Mn      Cr       O   SUM 
   472    .030    .001    .004    .009    .391    .000    .000    .000    .000    .000    .000    .000    .000    .000    .000    .436
   473    .024   -.001    .002    .003    .323    .000    .000    .000    .000    .000    .000    .000    .000    .000    .000    .352
   474    .013    .002   -.003    .009    .419    .000    .000    .000    .000    .000    .000    .000    .000    .000    .000    .439
   475    .021    .003   -.001    .001    .282    .000    .000    .000    .000    .000    .000    .000    .000    .000    .000    .306
   476    .011   -.002   -.001    .009    .242    .000    .000    .000    .000    .000    .000    .000    .000    .000    .000    .259
   477    .009    .000    .001    .004    .245    .000    .000    .000    .000    .000    .000    .000    .000    .000    .000    .259
   478    .007    .002    .003    .010    .268    .000    .000    .000    .000    .000    .000    .000    .000    .000    .000    .290
   479    .013    .002    .001    .006    .186    .000    .000    .000    .000    .000    .000    .000    .000    .000    .000    .207
   480    .003   -.002   -.002    .002    .129    .000    .000    .000    .000    .000    .000    .000    .000    .000    .000    .130
   481    .010   -.001    .003    .004    .291    .000    .000    .000    .000    .000    .000    .000    .000    .000    .000    .308
   482    .022    .000    .000    .008    .389    .000    .000    .000    .000    .000    .000    .000    .000    .000    .000    .420

AVER:     .015    .000    .001    .006    .288    .000    .000    .000    .000    .000    .000    .000    .000    .000    .000    .310
SDEV:     .008    .002    .002    .003    .089    .000    .000    .000    .000    .000    .000    .000    .000    .000    .000    .098
SERR:     .003    .000    .001    .001    .027    .000    .000    .000    .000    .000    .000    .000    .000    .000    .000
%RSD:    55.52  394.62  422.74   53.43   30.94     .00     .00     .00     .00     .00     .00     .00     .00     .00     .00
STDS:      257     541    1007     251      22     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---

STKF:    .4109   .9936   .5384   .4225   .5616     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---
STCT:    52.56  382.07  133.42   50.64   82.83     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---

UNKF:    .0001   .0000   .0000   .0000   .0029     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---
UNCT:      .02     .00     .00     .01     .42     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---
UNBG:      .08     .19     .31     .10     .07     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---

ZCOR:   1.1199  1.0725  1.0567  1.2265  1.0088     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---
KRAW:    .0003   .0000   .0000   .0001   .0051     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---
PKBG:     1.22    1.01    1.00    1.06    6.68     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---

Note that the total is around 0.3 wt%, mostly from the Ti measurement, therefore the matrix correction has to assume the matrix is almost pure Ti metal. And even if we calculate this as an oxide matrix, nothing changes except the Ti concentration by about 0.017 wt%:

ELEM:       Zr      Nb      La      Sr      Ti      Si      Ca      Al      Fe      Na       K      Mg      Mn      Cr       O
TYPE:     ANAL    ANAL    ANAL    ANAL    ANAL    SPEC    SPEC    SPEC    SPEC    SPEC    SPEC    SPEC    SPEC    SPEC    CALC
BGDS:     MULT    MULT    MULT    MULT    MULT
TIME:   240.00  240.00  240.00  240.00  240.00     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---
BEAM:    80.06   80.06   80.06   80.06   80.06     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---

ELEM:       Zr      Nb      La      Sr      Ti      Si      Ca      Al      Fe      Na       K      Mg      Mn      Cr       O   SUM 
   472    .031    .001    .004    .010    .414    .000    .000    .000    .000    .000    .000    .000    .000    .000    .290    .751
   473    .025   -.001    .002    .003    .343    .000    .000    .000    .000    .000    .000    .000    .000    .000    .238    .611
   474    .013    .002   -.003    .009    .445    .000    .000    .000    .000    .000    .000    .000    .000    .000    .304    .770
   475    .022    .003   -.001    .002    .299    .000    .000    .000    .000    .000    .000    .000    .000    .000    .209    .533
   476    .012   -.002   -.001    .009    .257    .000    .000    .000    .000    .000    .000    .000    .000    .000    .176    .450
   477    .009    .000    .001    .004    .260    .000    .000    .000    .000    .000    .000    .000    .000    .000    .178    .453
   478    .008    .002    .003    .010    .285    .000    .000    .000    .000    .000    .000    .000    .000    .000    .196    .503
   479    .013    .002    .001    .006    .197    .000    .000    .000    .000    .000    .000    .000    .000    .000    .138    .357
   480    .003   -.002   -.003    .002    .137    .000    .000    .000    .000    .000    .000    .000    .000    .000    .092    .231
   481    .010   -.001    .003    .004    .310    .000    .000    .000    .000    .000    .000    .000    .000    .000    .211    .538
   482    .023    .000    .001    .008    .413    .000    .000    .000    .000    .000    .000    .000    .000    .000    .286    .731

AVER:     .015    .000    .001    .006    .305    .000    .000    .000    .000    .000    .000    .000    .000    .000    .211    .539
SDEV:     .009    .002    .002    .003    .094    .000    .000    .000    .000    .000    .000    .000    .000    .000    .066    .169
SERR:     .003    .000    .001    .001    .028    .000    .000    .000    .000    .000    .000    .000    .000    .000    .020
%RSD:    55.61  394.67  423.76   51.36   30.90     .00     .00     .00     .00     .00     .00     .00     .00     .00   31.18
STDS:      257     541    1007     251      22     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---

STKF:    .4109   .9936   .5384   .4225   .5616     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---
STCT:    52.56  382.07  133.42   50.64   82.83     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---

UNKF:    .0001   .0000   .0000   .0001   .0029     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---
UNCT:      .02     .00     .00     .01     .42     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---
UNBG:      .08     .19     .31     .10     .07     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---

ZCOR:   1.1423  1.1009  1.1251  1.2484  1.0712     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---
KRAW:    .0003   .0000   .0000   .0001   .0051     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---
PKBG:     1.22    1.01    1.00    1.06    6.68     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---

Which we can see is from the change in the matrix correction for Ti Ka by about 6%. The other trace elements still show no change at all.  And what happens if we specify the correct matrix correctly as shown here:

Un   33 MA-1058 Rxn-1 (trav), Results in Elemental Weight Percents
 
ELEM:       Zr      Nb      La      Sr      Ti      Si      Ca      Al      Fe      Na       K      Mg      Mn      Cr       O
TYPE:     ANAL    ANAL    ANAL    ANAL    ANAL    SPEC    SPEC    SPEC    SPEC    SPEC    SPEC    SPEC    SPEC    SPEC    SPEC
BGDS:     MULT    MULT    MULT    MULT    MULT
TIME:   240.00  240.00  240.00  240.00  240.00     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---
BEAM:    80.06   80.06   80.06   80.06   80.06     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---

ELEM:       Zr      Nb      La      Sr      Ti      Si      Ca      Al      Fe      Na       K      Mg      Mn      Cr       O   SUM 
   472    .040    .002    .005    .011    .462  23.714  11.392   4.620   4.928    .994    .000  10.125    .108    .089  44.198 100.688
   473    .032   -.001    .002    .004    .383  23.714  11.392   4.620   4.928    .994    .000  10.125    .108    .089  44.198 100.589
   474    .017    .002   -.004    .011    .497  23.714  11.392   4.620   4.928    .994    .000  10.125    .108    .089  44.198 100.692
   475    .028    .004   -.002    .003    .334  23.714  11.392   4.620   4.928    .994    .000  10.125    .108    .089  44.198 100.535
   476    .015   -.003   -.001    .010    .287  23.714  11.392   4.620   4.928    .994    .000  10.125    .108    .089  44.198 100.476
   477    .012    .000    .001    .005    .291  23.714  11.392   4.620   4.928    .994    .000  10.125    .108    .089  44.198 100.478
   478    .010    .002    .003    .012    .319  23.714  11.392   4.620   4.928    .994    .000  10.125    .108    .089  44.198 100.513
   479    .017    .002    .001    .007    .220  23.714  11.392   4.620   4.928    .994    .000  10.125    .108    .089  44.198 100.415
   480    .004   -.002   -.003    .003    .154  23.714  11.392   4.620   4.928    .994    .000  10.125    .108    .089  44.198 100.324
   481    .013   -.001    .004    .005    .347  23.714  11.392   4.620   4.928    .994    .000  10.125    .108    .089  44.198 100.536
   482    .030    .000    .001    .009    .462  23.714  11.392   4.620   4.928    .994    .000  10.125    .108    .089  44.198 100.670

AVER:     .020    .001    .001    .007    .341  23.714  11.392   4.620   4.928    .994    .000  10.125    .108    .089  44.198 100.538
SDEV:     .011    .002    .003    .003    .105    .000    .000    .000    .000    .000    .000    .000    .000    .000    .000    .116
SERR:     .003    .001    .001    .001    .032    .000    .000    .000    .000    .000    .000    .000    .000    .000    .000
%RSD:    55.77  394.36  423.62   46.10   30.87     .00     .00     .00     .00     .00     .00     .00     .00     .00     .00
STDS:      257     541    1007     251      22     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---

STKF:    .4109   .9936   .5384   .4225   .5616     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---
STCT:    52.56  382.07  133.42   50.64   82.83     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---

UNKF:    .0001   .0000   .0000   .0001   .0029     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---
UNCT:      .02     .00     .00     .01     .42     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---
UNBG:      .08     .19     .31     .10     .07     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---

ZCOR:   1.4847  1.4029  1.2509  1.3181  1.1972     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---
KRAW:    .0003   .0000   .0000   .0001   .0051     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---
PKBG:     1.22    1.01    1.00    1.07    6.68     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---

Essentially no change in the trace elements, and only a rather small matrix effect on the Ti concentration. Or does it? Let try another matrix, say MgO:

Un   33 MA-1058 Rxn-1 (trav), Results in Elemental Weight Percents
 
ELEM:       Zr      Nb      La      Sr      Ti      Si      Ca      Al      Fe      Na       K      Mg      Mn      Cr       O
TYPE:     ANAL    ANAL    ANAL    ANAL    ANAL    SPEC    SPEC    SPEC    SPEC    SPEC    SPEC    SPEC    SPEC    SPEC    SPEC
BGDS:     MULT    MULT    MULT    MULT    MULT
TIME:   240.00  240.00  240.00  240.00  240.00     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---
BEAM:    80.06   80.06   80.06   80.06   80.06     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---

ELEM:       Zr      Nb      La      Sr      Ti      Si      Ca      Al      Fe      Na       K      Mg      Mn      Cr       O   SUM 
   472    .041    .002    .005    .020    .454    .000    .000    .000    .000    .000    .000  60.303    .000    .000  39.697 100.522
   473    .033   -.001    .002    .011    .377    .000    .000    .000    .000    .000    .000  60.303    .000    .000  39.697 100.422
   474    .018    .002   -.004    .019    .489    .000    .000    .000    .000    .000    .000  60.303    .000    .000  39.697 100.525
   475    .029    .004   -.002    .008    .328    .000    .000    .000    .000    .000    .000  60.303    .000    .000  39.697 100.368
   476    .015   -.003   -.001    .019    .282    .000    .000    .000    .000    .000    .000  60.303    .000    .000  39.697 100.312
   477    .012    .000    .001    .012    .286    .000    .000    .000    .000    .000    .000  60.303    .000    .000  39.697 100.312
   478    .010    .002    .003    .021    .313    .000    .000    .000    .000    .000    .000  60.303    .000    .000  39.697 100.349
   479    .017    .002    .001    .015    .216    .000    .000    .000    .000    .000    .000  60.303    .000    .000  39.697 100.252
   480    .004   -.002   -.003    .009    .151    .000    .000    .000    .000    .000    .000  60.303    .000    .000  39.697 100.160
   481    .014   -.001    .004    .012    .341    .000    .000    .000    .000    .000    .000  60.303    .000    .000  39.697 100.369
   482    .031    .000    .001    .018    .454    .000    .000    .000    .000    .000    .000  60.303    .000    .000  39.697 100.503

AVER:     .020    .001    .001    .015    .336    .000    .000    .000    .000    .000    .000  60.303    .000    .000  39.697 100.372
SDEV:     .011    .002    .003    .005    .104    .000    .000    .000    .000    .000    .000    .000    .000    .000    .000    .115
SERR:     .003    .001    .001    .001    .031    .000    .000    .000    .000    .000    .000    .000    .000    .000    .000
%RSD:    55.77  394.38  423.62   31.13   30.87     .00     .00     .00     .00     .00     .00     .00     .00     .00     .00
STDS:      257     541    1007     251      22     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---

STKF:    .4109   .9936   .5384   .4225   .5616     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---
STCT:    52.56  382.07  133.42   50.64   82.83     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---

UNKF:    .0001   .0000   .0000   .0001   .0029     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---
UNCT:      .02     .00     .00     .01     .42     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---
UNBG:      .08     .19     .31     .10     .07     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---

ZCOR:   1.5270  1.4383  1.2357  1.7923  1.1773     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---
KRAW:    .0003   .0000   .0000   .0002   .0051     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---
PKBG:     1.22    1.01    1.00    1.10    6.68     ---     ---     ---     ---     ---     ---     ---     ---     ---     ---

So we can see that even if we specify the matrix correction completely wrong, for trace elements, at least in this case, there is no significant effect and even for the minor element Ti, we see only a change of 0.025 wt%, or only about 2%.

What's the moral of this story?

1. Matrix corrections and standard composition accuracy is very important for major elements. These items should be your "focus".

2. For trace elements the accuracy of the background correction (and spectral interferences if present) are very important. Matrix corrections and standard composition accuracy, not so much if at all.

3. For minor elements, of course it's a "sliding scale" depending on the peak to background ratio (and the physics details), but if minor element accuracy is important we would want to have accurate background measurements primarily. The accuracy of the standard compositions and matrix correction effects are generally very small for minor elements.

To illustrate this last point here is the sample again calculated using all 10 matrix corrections in Probe for EPMA, where we can see the total variance of the Ti concentrations is quite small:

Summary of All Calculated (averaged) Matrix Corrections:
Un   33  MA-1058 Rxn-1 (trav)
LINEMU   Henke (LBL, 1985) < 10KeV / CITZMU > 10KeV

Elemental Weight Percents:
ELEM:       Zr      Nb      La      Sr      Ti      Si      Ca      Al      Fe      Na       K      Mg      Mn      Cr       O   TOTAL
     1    .020    .001    .001    .012    .336    .000    .000    .000    .000    .000    .000  60.303    .000    .000  39.697 100.370   Armstrong/Love Scott (default)
     2    .022    .001    .001    .015    .346    .000    .000    .000    .000    .000    .000  60.303    .000    .000  39.697 100.385   Conventional Philibert/Duncumb-Reed
     3    .021    .001    .001    .013    .335    .000    .000    .000    .000    .000    .000  60.303    .000    .000  39.697 100.370   Heinrich/Duncumb-Reed
     4    .022    .001    .001    .015    .337    .000    .000    .000    .000    .000    .000  60.303    .000    .000  39.697 100.374   Love-Scott I
     5    .021    .001    .001    .013    .336    .000    .000    .000    .000    .000    .000  60.303    .000    .000  39.697 100.372   Love-Scott II
     6    .023    .001    .001    .016    .348    .000    .000    .000    .000    .000    .000  60.303    .000    .000  39.697 100.388   Packwood Phi(pz) (EPQ-91)
     7    .021    .001    .001    .014    .334    .000    .000    .000    .000    .000    .000  60.303    .000    .000  39.697 100.371   Bastin (original) Phi(pz)
     8    .022    .001    .001    .015    .342    .000    .000    .000    .000    .000    .000  60.303    .000    .000  39.697 100.380   Bastin PROZA Phi(pz) (EPQ-91)
     9    .022    .001    .001    .015    .341    .000    .000    .000    .000    .000    .000  60.303    .000    .000  39.697 100.380   Pouchou and Pichoir-Full (PAP)
    10    .022    .001    .001    .014    .342    .000    .000    .000    .000    .000    .000  60.303    .000    .000  39.697 100.380   Pouchou and Pichoir-Simplified (XPP)

AVER:     .022    .001    .001    .014    .340    .000    .000    .000    .000    .000    .000  60.303    .000    .000  39.697 100.377
SDEV:     .001    .000    .000    .001    .005    .000    .000    .000    .000    .000    .000    .000    .000    .000    .000    .007
SERR:     .000    .000    .000    .000    .002    .000    .000    .000    .000    .000    .000    .000    .000    .000    .000

MIN:      .020    .001    .001    .012    .334    .000    .000    .000    .000    .000    .000  60.303    .000    .000  39.697 100.370
MAX:      .023    .001    .001    .016    .348    .000    .000    .000    .000    .000    .000  60.303    .000    .000  39.697 100.388

In fact the Ti variance for all 10 matrix corrections is only 50 PPM
« Last Edit: November 06, 2021, 04:17:19 PM by Probeman »
The only stupid question is the one not asked!

DavidAdams

  • Professor
  • ****
  • Posts: 54
Re: Specifying Unanalyzed Elements For a Proper Matrix Correction
« Reply #115 on: August 22, 2022, 05:51:18 PM »
Just FYI, we're just putting the finishing touches on a ferrous/ferric calculation for PFE (and CalcZAF!) that will calculate excess oxygen from ferric iron for minerals using the method of Droop (1987) *and* include it in the matrix correction.  This excess oxygen has a surprisingly large effect on the matrix correction physics, even for minerals such as ilmenite/magnetite.  I'm just posting this here so we'll have a link for the "interactive help" button in the Calculation Options dialog.

We hope to have a new version of both PFE and CalcZAF uploaded tonight or tomorrow, but first we want to give a big thanks to Andrew Locock, Anette von der Handt, John Fournelle and Emma Bullock, who provided the mineralogical expertise to allow us to implement this.

More details to follow soon but in the meantime the original paper by Droop is attached below

I'm wondering if someone much smarter than me on here can help me think through a problem related to Ferrous/Ferric ratios in glass. I have a series of basaltic glasses that were created by a researcher a long time ago under known and controlled fO2 conditions. A subset of each of these glasses were mounted in epoxy for microprobe analysis and a split of each sample was analysed using wet chemical methods to determine the Ferrous/Ferric content. I can't figure out how I can (if I can) use the relatively new Ferrous/Ferric in the software.

Any help would be greatly appreciated!!!
David Adams
The University of Auckland
Faculty of Science | School of Environment

Probeman

  • Emeritus
  • *****
  • Posts: 2771
  • Never sleeps...
    • John Donovan
Re: Specifying Unanalyzed Elements For a Proper Matrix Correction
« Reply #116 on: August 22, 2022, 06:59:16 PM »
Short answer is you can't for glasses.

The ferric/ferrous excess oxygen calculation is based on charge balance (Droop, 1987) so a formula (number of cations and oxygens) is required for it to work. We even got it to work for most amphibole compositions with a lot of help from Andrew Locock and Aurelien Moy.

But I guess I'm not sure why you want to try this when you already have ferric/ferrous contents from wet chemistry.  Just specify the excess oxygen from ferric iron in PFE and it will perform a matrix correction and should give you nicer totals.
« Last Edit: August 22, 2022, 07:24:22 PM by Probeman »
The only stupid question is the one not asked!

AndrewLocock

  • Professor
  • ****
  • Posts: 98
    • University of Alberta Electron Microprobe Laboratory
Re: Specifying Unanalyzed Elements For a Proper Matrix Correction
« Reply #117 on: August 23, 2022, 08:17:38 AM »
The following conversions for Fe2O3, FeO and excess O may be useful to some.

Conversions between FeO and Fe2O3 in wt%:   
   Fe2O3 = FeO * (159.6882 / [2 * 71.8444]) = FeO * 1.111348
   FeO = Fe2O3 * ([2 * 71.8444] / 159.6882) = Fe2O3 * 0.8998085

Conversions for Fe3+-bearing compounds with "excess oxygen":
   Fe2O3 wt% = wt% O "excess oxygen"* (molar mass Fe2O3/molar mass O)
   Fe2O3 wt% = wt% O "excess oxygen" * (159.6882/15.9994)
   Fe2O3 wt% = wt% O "excess oxygen" * 9.980887

Example: hematite, with 89.981 wt% FeOtotal and "excess oxygen" 10.019 wt% O:
   Fe2O3 wt% = 10.019 wt% O * 9.980887 = 100.00
   FeOfinal = FeOtotal – Fe2O3 * 0.8998085 = 0.
   Thus, hematite is 100.00 wt% Fe2O3

Example: magnetite, with 93.09 wt% FeOtotal and "excess oxygen" 6.91 wt% O:
   Fe2O3 wt% = 6.91 wt% O * 9.980887 = 68.97
   FeOfinal = FeOtotal – Fe2O3 * 0.8998085 = 93.09 – 62.06 = 31.06
   Thus, magnetite has 68.97 Fe2O3 and 31.06 FeO, sum 100.00 wt%

Probeman

  • Emeritus
  • *****
  • Posts: 2771
  • Never sleeps...
    • John Donovan
Re: Specifying Unanalyzed Elements For a Proper Matrix Correction
« Reply #118 on: August 23, 2022, 08:52:05 AM »
Just FYI, we're just putting the finishing touches on a ferrous/ferric calculation for PFE (and CalcZAF!) that will calculate excess oxygen from ferric iron for minerals using the method of Droop (1987) *and* include it in the matrix correction.  This excess oxygen has a surprisingly large effect on the matrix correction physics, even for minerals such as ilmenite/magnetite.  I'm just posting this here so we'll have a link for the "interactive help" button in the Calculation Options dialog.

We hope to have a new version of both PFE and CalcZAF uploaded tonight or tomorrow, but first we want to give a big thanks to Andrew Locock, Anette von der Handt, John Fournelle and Emma Bullock, who provided the mineralogical expertise to allow us to implement this.

More details to follow soon but in the meantime the original paper by Droop is attached below

I'm wondering if someone much smarter than me on here can help me think through a problem related to Ferrous/Ferric ratios in glass. I have a series of basaltic glasses that were created by a researcher a long time ago under known and controlled fO2 conditions. A subset of each of these glasses were mounted in epoxy for microprobe analysis and a split of each sample was analysed using wet chemical methods to determine the Ferrous/Ferric content. I can't figure out how I can (if I can) use the relatively new Ferrous/Ferric in the software.

Any help would be greatly appreciated!!!

Actually I just remembered that there there is a method you might be able to use to determine ferric/ferrous ratios in glasses, but it's non-trivial.

That is the method of water by difference (by analyzing oxygen) first proposed by Barbara Nash at University of Utah.  Basically one analyzes for oxygen along with all the cations and then subtracts the oxygen from cations from the analyzed oxygen.

I made a study of this method using the features available in Probe for EPMA (curved backgrounds, Area Peak Factors, empirical mass absorption coefficients, including water in the matrix correction, etc.) on some synthetic glasses from Tony Withers (Universität Bayreuth) and the results are discussed here :

https://probesoftware.com/smf/index.php?topic=922.0

and the details are in the attached pdf at the above link.  The point is that in Probe for EPMA one can assign the excess oxygen as water or OH to be included in the matrix correction, or one can simply assume the excess oxygen is due to ferric iron. In all cases some assumptions have to be made.

It was pretty difficult to get enough accuracy with oxygen to obtain an accurate H2O value (but I did analyze them "blind" in the first effort and they came out very close to the FTIR values), and I think it would be even harder to get the excess oxygen from ferric iron, but it might be possible.
« Last Edit: August 23, 2022, 08:55:17 AM by Probeman »
The only stupid question is the one not asked!

DavidAdams

  • Professor
  • ****
  • Posts: 54
Re: Specifying Unanalyzed Elements For a Proper Matrix Correction
« Reply #119 on: August 23, 2022, 02:55:11 PM »
The following conversions for Fe2O3, FeO and excess O may be useful to some.

Conversions between FeO and Fe2O3 in wt%:   
   Fe2O3 = FeO * (159.6882 / [2 * 71.8444]) = FeO * 1.111348
   FeO = Fe2O3 * ([2 * 71.8444] / 159.6882) = Fe2O3 * 0.8998085

Conversions for Fe3+-bearing compounds with "excess oxygen":
   Fe2O3 wt% = wt% O "excess oxygen"* (molar mass Fe2O3/molar mass O)
   Fe2O3 wt% = wt% O "excess oxygen" * (159.6882/15.9994)
   Fe2O3 wt% = wt% O "excess oxygen" * 9.980887

Example: hematite, with 89.981 wt% FeOtotal and "excess oxygen" 10.019 wt% O:
   Fe2O3 wt% = 10.019 wt% O * 9.980887 = 100.00
   FeOfinal = FeOtotal – Fe2O3 * 0.8998085 = 0.
   Thus, hematite is 100.00 wt% Fe2O3

Example: magnetite, with 93.09 wt% FeOtotal and "excess oxygen" 6.91 wt% O:
   Fe2O3 wt% = 6.91 wt% O * 9.980887 = 68.97
   FeOfinal = FeOtotal – Fe2O3 * 0.8998085 = 93.09 – 62.06 = 31.06
   Thus, magnetite has 68.97 Fe2O3 and 31.06 FeO, sum 100.00 wt%

Thanks, Andrew! That's really helpful! :)
David Adams
The University of Auckland
Faculty of Science | School of Environment