Author Topic: Specifying Unanalyzed Elements For a Proper Matrix Correction (Read 47977 times)

John Donovan · « **Reply #45 on:** November 28, 2015, 07:43:34 AM »

Hi Andrew,
This is very useful stuff. Thank-you.
john

John Donovan · « **Reply #46 on:** December 09, 2015, 12:52:21 PM »

I should mention that Julien Allaz's mineral re-normalization code has been translated from php to VB, so I will be able to implement this in Probe for EPMA at some point "real soon now"...

AndrewLocock · « **Reply #47 on:** January 19, 2016, 10:06:32 AM »

Here is an updated version of the Excel spreadsheet, which now properly deals with cases where both halogens and hydrogen are present in oxide materials.

John Donovan · « **Reply #48 on:** July 06, 2018, 03:15:37 PM »

As many of you already know, there are two ways to specify unanalyzed elements *by difference* in the Probe for EPMA matrix corrections. Here is a screen shot of the Calculation Options dialog in PFE showing the two "difference" methods:

So one could specify OH by difference by selecting the "element by difference" option, then selecting hydrogen as the element by difference, then making sure that the oxygen stoichiometry for hydrogen is changed from the default of 1 cation and 2 oxygens, to 1 hydrogen and 1 oxygen, in the Elements/Cations dialog.

Then the program will calculate OH by difference. The other method is to use the "formula by difference" method where one specifies the unanalyzed elements as a formula, in this example, as OH, and the program adds them in automatically during the matrix calculation. The benefit of including these unanalyzed elements by difference into the matrix correction, is that the presence (or absence) of the unanalyzed elements can affect the concentrations of the analyzed elements. Thus affecting the calculation by difference. This is particularly import for so called "water by difference" measurements as discussed here:

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

As another example of elements by difference, one might be measuring traces in a matrix and have no need to measure the major elements. Say, traces in quartz or zircon, where one would measure the trace elements and merely specify the matrix elements as SiO2 or ZrSiO4 respectively, as a formula by difference. This is discussed here:

http://probesoftware.com/smf/index.php?topic=42.msg2568#msg2568

This formula by difference methods has worked well for when calculating without stoichiomeric oxygen, but after Gareth Seward and Marisa Acosta reported to us that this formula by difference method was not working correctly when oxygen was being calculated by stoichiomtry, we took a look and *finally* managed to find the problem and fixed the code. This latest version of PFE will fix this issue.

Here is an example of an apatite sample where the students wanted to calculate OH by difference and calculate oxygen by stoichiometry:

Un 52 BMA-062-16fa_g25, Results in Elemental Weight Percents SPEC: O H TYPE: FORM FORM AVER: 38.302 .038 SDEV: .504 .041 ELEM: Ca P S Al Fe As Ba K BGDS: LIN EXP LIN LIN LIN EXP LIN LIN TIME: 25.00 8.00 25.00 25.00 30.00 60.00 90.00 90.00 BEAM: 30.06 30.06 30.06 30.06 30.06 99.04 99.04 99.04 ELEM: Ca P S Al Fe As Ba K SUM 384 38.598 18.351 .032 .007 .181 -.076 -.008 .022 100.000 385 38.462 18.005 .019 -.002 .167 -.003 -.001 .013 100.000 386 38.210 18.037 .018 .002 .183 .030 -.007 .010 100.000 387 38.235 17.974 .032 .008 .194 -.010 -.006 .003 100.000 388 38.602 18.336 .037 -.007 .181 .030 -.004 .002 100.000 389 37.771 18.007 .030 .011 .218 .074 .001 .059 100.000 AVER: 38.313 18.118 .028 .003 .187 .007 -.004 .018 100.000 SDEV: .315 .175 .008 .007 .017 .051 .004 .021 .000 SERR: .129 .072 .003 .003 .007 .021 .001 .009 %RSD: .82 .97 28.02 213.60 9.28 676.66 -86.58 116.07 STDS: 285 285 327 336 160 662 835 336 STKF: .3596 .1601 .2210 .1333 .0654 .5052 .7431 .0409 STCT: 738.0 4696.5 1704.0 6059.0 1413.2 583.1 3099.0 588.4 UNKF: .3609 .1598 .0002 .0000 .0016 .0001 .0000 .0002 UNCT: 740.6 4689.7 1.9 1.1 34.1 .1 -.1 2.7 UNBG: 6.8 10.7 2.9 30.0 27.0 12.9 6.5 10.4 ZCOR: 1.0615 1.1335 1.1605 1.3622 1.1852 1.3270 1.4078 .9827 KRAW: 1.0036 .9986 .0011 .0002 .0242 .0001 .0000 .0045 PKBG: 110.25 468.43 1.67 1.04 2.27 1.01 .98 1.26 INT%: ---- ---- ---- ---- ---- ---- ---- ---- TDI%: .046 .488 4.073 1.487 2.693 ---- ---- ---- DEV%: .4 .2 24.6 2.3 1.6 ---- ---- ---- TDIF: LOG-LIN LOG-LIN LOG-LIN LOG-LIN LOG-LIN ---- ---- ---- TDIT: 56.67 37.83 55.50 55.83 57.83 ---- ---- ---- TDII: 747. 4700. 4.56 31.2 61.7 ---- ---- ---- TDIL: 6.62 8.46 1.52 3.44 4.12 ---- ---- ---- ELEM: Mn Na Sr F Sm Nd Cl Y BGDS: LIN LIN LIN LIN LIN LIN LIN LIN TIME: 60.00 90.00 60.00 60.00 60.00 90.00 40.00 90.00 BEAM: 99.04 99.04 99.04 99.04 99.04 99.04 99.04 99.04 ELEM: Mn Na Sr F Sm Nd Cl Y SUM 384 .302 .115 .050 2.828 .067 .102 .125 .337 100.000 385 .294 .118 .049 3.490 .027 .115 .132 .297 100.000 386 .332 .102 .043 3.229 .024 .085 .147 .257 100.000 387 .361 .102 .057 3.950 .047 .055 .092 .229 100.000 388 .309 .119 .041 3.970 .022 .091 .115 .295 100.000 389 .313 .124 .042 4.081 .007 .126 .133 .370 100.000 AVER: .318 .113 .047 3.591 .032 .096 .124 .298 100.000 SDEV: .024 .009 .006 .497 .021 .025 .019 .051 .000 SERR: .010 .004 .003 .203 .009 .010 .008 .021 %RSD: 7.63 8.01 13.19 13.84 65.51 26.11 15.42 17.23 STDS: 25 336 251 835 1011 1009 285 1016 STKF: .7341 .0735 .4267 .1715 .5260 .5221 .0602 .4481 STCT: 4303.0 1786.2 3853.8 561.2 1133.9 3358.0 836.0 1076.4 UNKF: .0026 .0005 .0004 .0071 .0002 .0007 .0011 .0026 UNCT: 15.4 12.7 3.6 23.1 .5 4.4 15.1 6.2 UNBG: 5.8 10.3 5.5 2.3 4.8 11.8 9.0 2.5 ZCOR: 1.2109 2.1625 1.1789 5.0816 1.4239 1.3990 1.1400 1.1620 KRAW: .0036 .0071 .0009 .0412 .0004 .0013 .0181 .0057 PKBG: 3.67 2.24 1.66 11.03 1.10 1.37 2.68 3.45 INT%: ---- ---- ---- -21.28 -6.97 -.95 -.31 ---- TDI%: ---- ---- ---- ---- ---- ---- ---- ---- DEV%: ---- ---- ---- ---- ---- ---- ---- ---- TDIF: ---- ---- ---- ---- ---- ---- ---- ---- TDIT: ---- ---- ---- ---- ---- ---- ---- ---- TDII: ---- ---- ---- ---- ---- ---- ---- ---- TDIL: ---- ---- ---- ---- ---- ---- ---- ---- ELEM: Eu Gd Mg La Ce Si BGDS: AVG LIN LIN LIN LIN LIN TIME: 60.00 60.00 90.00 60.00 60.00 60.00 BEAM: 99.04 99.04 99.04 99.04 99.04 99.04 ELEM: Eu Gd Mg La Ce Si SUM 384 .029 .041 .045 .025 .133 .099 100.000 385 -.005 .057 .044 .015 .112 .090 100.000 386 -.008 .047 .039 .017 .097 .075 100.000 387 .015 .029 .036 .005 .079 .058 100.000 388 .024 .051 .041 .016 .129 .094 100.000 389 .016 .056 .124 .023 .176 .295 100.000 AVER: .012 .047 .055 .017 .121 .118 100.000 SDEV: .015 .011 .034 .007 .033 .088 .000 SERR: .006 .004 .014 .003 .014 .036 %RSD: 130.06 22.47 61.89 42.93 27.58 73.94 STDS: 1004 1005 160 1007 1001 160 STKF: .5289 .5281 .0776 .5143 .5111 .1621 STCT: 1239.6 4995.6 3016.8 11742.1 3550.7 5699.0 UNKF: .0001 .0003 .0003 .0001 .0009 .0010 UNCT: .2 3.0 13.4 2.7 6.0 35.8 UNBG: 5.6 22.3 19.3 63.8 25.3 5.8 ZCOR: 1.4453 1.4722 1.5909 1.3981 1.3935 1.1629 KRAW: .0002 .0006 .0044 .0002 .0017 .0063 PKBG: 1.04 1.14 1.69 1.04 1.24 7.09 INT%: 11.30 -17.82 ---- -5.08 -.03 ---- TDI%: ---- ---- ---- ---- ---- ---- DEV%: ---- ---- ---- ---- ---- ---- TDIF: ---- ---- ---- ---- ---- ---- TDIT: ---- ---- ---- ---- ---- ---- TDII: ---- ---- ---- ---- ---- ---- TDIL: ---- ---- ---- ---- ---- ---- Un 52 BMA-062-16fa_g25, Results in Oxide Weight Percents SPEC: O HO TYPE: FORM FORM AVER: -1.540 .638 SDEV: .207 .685 ELEM: CaO P2O5 SO3 Al2O3 FeO As2O3 BaO K2O SUM 384 54.007 42.049 .081 .014 .232 -.100 -.008 .026 100.000 385 53.816 41.256 .048 -.004 .215 -.004 -.001 .016 100.000 386 53.464 41.330 .044 .005 .235 .040 -.008 .013 100.000 387 53.498 41.185 .081 .015 .249 -.014 -.007 .004 100.000 388 54.011 42.014 .093 -.013 .233 .040 -.004 .003 100.000 389 52.849 41.262 .075 .020 .280 .097 .001 .071 100.000 AVER: 53.608 41.516 .070 .006 .241 .010 -.005 .022 100.000 SDEV: .441 .402 .020 .013 .022 .067 .004 .025 .000 SERR: .180 .164 .008 .005 .009 .027 .002 .010 %RSD: .82 .97 28.02 213.60 9.28 676.66 -86.58 116.07 STDS: 285 285 327 336 160 662 835 336 ELEM: MnO Na2O SrO F Sm2O3 Nd2O3 Cl Y2O3 SUM 384 .390 .155 .059 2.828 .077 .119 .125 .429 100.000 385 .380 .158 .057 3.490 .031 .134 .132 .377 100.000 386 .429 .138 .051 3.229 .028 .100 .147 .327 100.000 387 .466 .138 .068 3.950 .054 .064 .092 .291 100.000 388 .398 .161 .048 3.970 .025 .106 .115 .374 100.000 389 .404 .167 .050 4.081 .008 .147 .133 .470 100.000 AVER: .411 .153 .055 3.591 .037 .112 .124 .378 100.000 SDEV: .031 .012 .007 .497 .025 .029 .019 .065 .000 SERR: .013 .005 .003 .203 .010 .012 .008 .027 %RSD: 7.63 8.01 13.19 13.84 65.51 26.11 15.42 17.23 STDS: 25 336 251 835 1011 1009 285 1016 ELEM: Eu2O3 Gd2O3 MgO La2O3 Ce2O3 SiO2 SUM 384 .033 .047 .075 .029 .155 .213 100.000 385 -.006 .066 .073 .017 .131 .192 100.000 386 -.010 .054 .065 .020 .114 .160 100.000 387 .018 .034 .059 .005 .093 .124 100.000 388 .028 .059 .069 .019 .151 .202 100.000 389 .019 .064 .205 .027 .206 .630 100.000 AVER: .014 .054 .091 .020 .142 .254 100.000 SDEV: .018 .012 .056 .008 .039 .187 .000 SERR: .007 .005 .023 .003 .016 .077 %RSD: 130.06 22.47 61.89 42.93 27.58 73.94 STDS: 1004 1005 160 1007 1001 160
Notice that the oxygen concentration in the oxide output is negative. This is because the program is also performing the correction for the halogen equivalent for oxygen in the matrix calculation! Pretty cool!

If you'd like to learn more about the halogen equivalence for stoichiometric oxygen calculation feature see this post:

http://probesoftware.com/smf/index.php?topic=8.msg1127#msg1127

AndrewLocock · « **Reply #49 on:** September 28, 2018, 09:10:25 AM »

How does one delete previously-added unanalyzed elements from the list of elements?

Sometimes, one might want to try adding an unanalyzed element to set of analyses, as a test.
However, one may not want to keep such elements.
It would be nice to easily delete them, if one doesn't want them.
Thanks,
Andrew

John Donovan · « **Reply #50 on:** September 28, 2018, 11:35:58 AM »

Quote from: AndrewLocock on September 28, 2018, 09:10:25 AM

How does one delete previously-added unanalyzed elements from the list of elements?

Sometimes, one might want to try adding an unanalyzed element to set of analyses, as a test.
However, one may not want to keep such elements.
It would be nice to easily delete them, if one doesn't want them.

Hi Andrew,
It gets complicated because it's difficult to know what an element might be utilized for, so to be safe we only allow elements to be deleted from the Acquire! Elements/Cations dialog. Sorry.
john

Probeman · « **Reply #51 on:** October 12, 2018, 03:37:43 PM »

This topic is entitled "Specifying Unanalyzed Elements For a Proper Matrix Correction", but it could just as well have been titled "Including unanalyzed elements for proper quantification of analyzed elements".

I recently responded to a question on the JEOL Listserver about how to add the carbonate molecule to a carbonate analysis using the JEOL software. I don't know the JEOL software, but then it was suggested performing the calculation off-line perhaps using Excel. This is *not* a good idea as most of you know, because the matrix correction will be significantly wrong and the quantification of the analyzed elements will be significantly wrong.

If the total of the elements in the matrix correction is significantly under 100%, the matrix correction physics will be significantly wrong in most cases. For example carbonates.

Here is an analysis of a standard dolomite where because the software knows that it is a standard, it can automatically specify the unanalyzed elements to get a correct quantification for the elements that were acquired:

St 141 Set 1 Dolomite (Harvard #105064), Results in Elemental Weight Percents ELEM: Ca Mn P Mg Fe C O TYPE: ANAL ANAL ANAL ANAL ANAL SPEC SPEC BGDS: LIN LIN LIN LIN LIN TIME: 10.00 10.00 10.00 10.00 10.00 --- --- BEAM: 30.01 30.01 30.01 30.01 30.01 --- --- ELEM: Ca Mn P Mg Fe C O SUM 43 21.796 .081 -.040 13.417 .145 12.986 52.021 100.406 44 21.698 .036 .025 13.332 .145 12.986 52.021 100.243 45 21.876 .105 .024 13.396 .168 12.986 52.021 100.576 46 22.000 .068 -.014 13.224 .145 12.986 52.021 100.430 AVER: 21.843 .073 -.001 13.342 .151 12.986 52.021 100.414 SDEV: .128 .029 .032 .087 .011 .000 .000 .136 SERR: .064 .014 .016 .043 .006 .000 .000 %RSD: .58 39.60-2300.07 .65 7.43 .00 .00 PUBL: 21.841 .015 n.a. 13.195 .062 12.986 52.021 100.120 %VAR: .01 383.93 --- 1.12 143.23 .00 .00 DIFF: .001 .058 --- .147 .089 .000 .000 STDS: 138 140 285 139 145 --- --- STKF: .3789 .3969 .1600 .1957 .4258 --- --- STCT: 128.18 134.89 56.49 65.78 142.56 --- --- UNKF: .2030 .0006 .0000 .0853 .0012 --- --- UNCT: 68.66 .20 .00 28.69 .42 --- --- UNBG: .14 .77 1.00 .06 .49 --- --- ZCOR: 1.0763 1.2369 1.2287 1.5634 1.2161 --- --- KRAW: .5356 .0015 -.0001 .4362 .0029 --- --- PKBG: 484.48 1.27 1.00 473.26 1.85 --- ---
Note that carbon and oxygen were automatically loaded from the standard database into Probe for EPMA, because the software knows that it is a standard. But what if this was an unknown carbonate? Here is the same material (dolomite standard), but this time acquired as an unknown:

Un 6 Dolomite (Harvard #105064), Results in Elemental Weight Percents ELEM: Ca Mn P Mg Fe C TYPE: ANAL ANAL ANAL ANAL ANAL SPEC BGDS: LIN LIN LIN LIN LIN TIME: 10.00 10.00 10.00 10.00 10.00 --- BEAM: 29.99 29.99 29.99 29.99 29.99 --- ELEM: Ca Mn P Mg Fe C SUM 47 21.220 .033 -.006 11.758 .135 .000 33.140 48 21.136 .021 .010 11.939 .150 .000 33.258 49 21.296 .055 .013 12.068 .164 .000 33.596 50 21.292 .080 .067 12.059 .128 .000 33.626 51 21.237 .066 .026 11.925 .117 .000 33.371 52 21.278 .090 .017 11.796 .156 .000 33.337 53 20.925 .121 .084 11.887 .161 .000 33.178 AVER: 21.198 .067 .030 11.919 .145 .000 33.358 SDEV: .132 .034 .033 .119 .018 .000 .191 SERR: .050 .013 .012 .045 .007 .000 %RSD: .62 51.24 108.72 1.00 12.32 .00 STDS: 138 140 285 139 145 --- STKF: .3789 .3969 .1600 .1957 .4258 --- STCT: 128.18 134.89 56.49 65.78 142.56 --- UNKF: .2047 .0006 .0002 .0852 .0012 --- UNCT: 69.26 .19 .08 28.63 .41 --- UNBG: .13 .78 .97 .05 .51 --- ZCOR: 1.0354 1.2003 1.2679 1.3995 1.1694 --- KRAW: .5403 .0014 .0015 .4353 .0029 --- PKBG: 529.78 1.25 1.09 606.24 1.81 ---
First note the very low total (~33 wt.%). Because the software does not know that this is a carbonate, it cannot automatically specify the unanalyzed carbon and oxygen. And note that although Ca (Ka) isn't so affected (21.2 wt.% in the unknown vs. 21.8 wt.% in the standard), the Mg (Ka) is very much affected (11.9 wt.% in the unknown vs. 13.3 wt.% in the standard!).

Observe the differences in the magnitude of the matrix corrections for Ca (Ka) and Mg (Ka), in the line labeled ZCOR, and you will understand why.

So how to add in the missing carbonate molecule? For the JEOL software Pete McSwiggen's suggestion is good because I assume that the JEOL software is including the specified unanalyzed elements in the matrix correction.

In Probe for EPMA one can do the same thing by first adding carbon as a unanalyzed element in the Elements/Cations dialog, then simply checking the Calculate with Stoichiometric Oxygen option, and finally specify 0.333 atoms of carbon per oxygen molecule as seen here:

Once that is done, the analysis proceeds as follows:

Un 6 Dolomite (Harvard #105064), Results in Elemental Weight Percents ELEM: Ca Mn P Mg Fe C O TYPE: ANAL ANAL ANAL ANAL ANAL STOI CALC BGDS: LIN LIN LIN LIN LIN TIME: 10.00 10.00 10.00 10.00 10.00 --- --- BEAM: 29.99 29.99 29.99 29.99 29.99 --- --- ELEM: Ca Mn P Mg Fe C O SUM 47 22.060 .034 -.006 13.130 .141 12.959 52.019 100.336 48 21.965 .022 .010 13.349 .156 12.955 52.136 100.594 49 22.130 .057 .012 13.486 .171 12.925 52.228 101.008 50 22.123 .082 .065 13.482 .133 12.929 52.298 101.113 51 22.071 .068 .025 13.327 .122 12.945 52.160 100.719 52 22.122 .092 .016 13.162 .162 12.942 52.071 100.567 53 21.748 .125 .081 13.286 .167 12.970 52.172 100.549 AVER: 22.031 .069 .029 13.318 .150 12.946 52.155 100.698 SDEV: .137 .035 .032 .140 .019 .016 .093 .274 SERR: .052 .013 .012 .053 .007 .006 .035 %RSD: .62 51.26 108.66 1.05 12.33 .13 .18 STDS: 138 140 285 139 145 --- --- STKF: .3789 .3969 .1600 .1957 .4258 --- --- STCT: 128.18 134.89 56.49 65.78 142.56 --- --- UNKF: .2047 .0006 .0002 .0852 .0012 --- --- UNCT: 69.26 .19 .08 28.63 .41 --- --- UNBG: .13 .78 .97 .05 .51 --- --- ZCOR: 1.0761 1.2369 1.2276 1.5637 1.2160 --- --- KRAW: .5403 .0014 .0015 .4353 .0029 --- --- PKBG: 529.78 1.25 1.09 606.24 1.81 --- ---
Note that now the concentrations and matrix corrections for Ca and Mg are correct.

By the way, the same calculation options will work for any carbonate such as calcite, magnesite, siderite, etc.

Probeman · « **Reply #52 on:** October 13, 2018, 03:01:57 PM »

Since I mentioned the effect of not including water by difference as it affects the matrix correction in my JEOL Listserver post, I will "show my work" as to how the various unanalyzed elements affects the analyzed elements, as they are added to the matrix correction of a hydrous glass. For a publication reference see this paper:

https://link.springer.com/article/10.1007%2Fs00445-005-0003-z

In this example I selected an experimental glass from Withers, which nominally contains 4.9 wt.% H2O, but FTIR showed 5.06 wt%. Here is the analysis of this glass as an unknown with nothing but the analyzed elements:

Un 17 Withers-N5, Results in Elemental Weight Percents ELEM: Na K Cl Ba F Ti Fe Mn Ca Si Al Mg O H O TYPE: ANAL ANAL ANAL ANAL ANAL ANAL ANAL ANAL ANAL ANAL ANAL ANAL ANAL SPEC SPEC BGDS: MAN LIN LIN LIN LIN LOW MAN LIN MAN MAN MAN MAN EXP TIME: 60.00 20.00 10.00 20.00 40.00 10.00 40.00 10.00 20.00 20.00 20.00 60.00 --- --- --- BEAM: 9.98 9.98 9.98 9.98 9.98 9.98 9.98 9.98 9.98 9.98 9.98 9.98 --- --- --- ELEM: Na K Cl Ba F Ti Fe Mn Ca Si Al Mg O-D H O SUM 574 1.206 3.594 .218 .029 .046 .106 2.963 .058 .154 30.812 4.487 -.010 --- .000 .000 43.664 575 1.187 3.590 .221 .016 .045 .122 2.951 .076 .140 30.681 4.454 -.014 --- .000 .000 43.469 576 1.163 3.469 .192 .032 .054 .135 2.960 .080 .121 30.677 4.413 -.019 --- .000 .000 43.275 577 1.171 3.500 .276 .040 .059 .096 2.993 .056 .109 30.817 4.512 -.020 --- .000 .000 43.610 578 1.225 3.549 .276 -.013 .045 .071 3.001 .078 .108 30.727 4.451 -.012 --- .000 .000 43.505 579 1.218 3.559 .255 -.077 .038 .122 3.019 .079 .123 30.718 4.502 -.015 --- .000 .000 43.541 580 1.143 3.579 .238 .015 .048 .100 2.954 .052 .124 30.733 4.422 -.020 --- .000 .000 43.387 581 1.235 3.438 .264 -.019 .065 .116 2.933 .048 .144 30.823 4.443 -.015 --- .000 .000 43.475 582 1.170 3.501 .175 -.025 .043 .132 3.028 .049 .112 30.879 4.462 -.012 --- .000 .000 43.514 583 1.160 3.556 .233 -.061 .018 .145 2.995 .030 .136 30.874 4.465 -.017 --- .000 .000 43.534 584 1.183 3.491 .227 .004 .018 .122 2.956 .061 .125 30.945 4.416 -.017 --- .000 .000 43.532 585 1.148 3.503 .204 .029 .038 .119 2.978 .059 .153 30.943 4.465 -.015 --- .000 .000 43.623 AVER: 1.184 3.527 .232 -.002 .043 .115 2.978 .060 .129 30.802 4.458 -.016 --- .000 .000 43.511 SDEV: .031 .051 .032 .038 .014 .020 .029 .015 .016 .095 .032 .003 --- .000 .000 .105 SERR: .009 .015 .009 .011 .004 .006 .009 .004 .005 .028 .009 .001 --- .000 .000 %RSD: 2.59 1.43 13.87-1520.63 33.07 17.23 .99 25.47 12.47 .31 .72 -21.27 --- .00 .00 STDS: 336 374 285 835 835 22 395 25 358 162 336 12 --- --- --- STKF: .0735 .1132 .0602 .7431 .1715 .5547 .6779 .7341 .1693 .2018 .1333 .4736 --- --- --- STCT: 2447.9 2423.7 839.3 8520.6 2398.7 6097.6 14136.8 13590.2 2247.6 34290.6 23223.7 24410.3 --- --- --- UNKF: .0082 .0301 .0017 .0000 .0002 .0010 .0258 .0005 .0011 .2710 .0411 -.0001 --- --- --- UNCT: 274.6 645.2 24.2 -.2 2.7 10.8 538.5 9.5 15.2 46037.3 7157.6 -6.4 --- --- --- UNBG: 16.0 12.6 5.0 29.0 3.8 5.8 30.1 16.1 5.4 173.0 135.8 23.3 --- --- --- ZCOR: 1.4355 1.1708 1.3315 1.3476 2.1906 1.1747 1.1530 1.1748 1.1290 1.1368 1.0853 1.2443 --- --- --- KRAW: .1122 .2662 .0289 .0000 .0011 .0018 .0381 .0007 .0068 1.3426 .3082 -.0003 --- --- --- PKBG: 18.14 52.35 6.39 1.00 1.77 3.03 18.90 1.61 3.79 267.17 53.70 .72 --- --- --- INT%: ---- ---- ---- -3.80 ---- -.02 ---- ---- ---- ---- ---- ---- --- --- ---
If we calculate the above elemental analysis as oxide formulas (but still without oxygen included in the matrix correction), we get the following results:

Un 17 Withers-N5, Results in Oxide Weight Percents ELEM: Na2O K2O Cl BaO F TiO2 FeO MnO CaO SiO2 Al2O3 MgO O-D H2O O SUM 574 1.625 4.330 .218 .033 .046 .177 3.812 .075 .216 65.918 8.479 -.016 --- .000 .000 84.911 575 1.600 4.324 .221 .018 .045 .203 3.797 .098 .196 65.639 8.415 -.024 --- .000 .000 84.533 576 1.568 4.179 .192 .035 .054 .225 3.808 .103 .169 65.629 8.338 -.032 --- .000 .000 84.267 577 1.578 4.216 .276 .045 .059 .161 3.851 .072 .153 65.928 8.526 -.033 --- .000 .000 84.831 578 1.651 4.276 .276 -.015 .045 .118 3.861 .100 .152 65.736 8.410 -.020 --- .000 .000 84.588 579 1.642 4.287 .255 -.086 .038 .204 3.884 .102 .171 65.716 8.507 -.025 --- .000 .000 84.696 580 1.540 4.311 .238 .017 .048 .166 3.800 .068 .173 65.749 8.356 -.034 --- .000 .000 84.432 581 1.665 4.141 .264 -.021 .065 .193 3.773 .062 .202 65.942 8.395 -.025 --- .000 .000 84.656 582 1.577 4.217 .175 -.028 .043 .220 3.895 .063 .157 66.061 8.431 -.020 --- .000 .000 84.792 583 1.564 4.283 .233 -.068 .018 .241 3.853 .039 .190 66.051 8.437 -.028 --- .000 .000 84.813 584 1.595 4.205 .227 .004 .018 .204 3.803 .079 .175 66.202 8.344 -.027 --- .000 .000 84.829 585 1.547 4.220 .204 .032 .038 .198 3.831 .076 .214 66.199 8.436 -.024 --- .000 .000 84.971 AVER: 1.596 4.249 .232 -.003 .043 .192 3.831 .078 .181 65.897 8.423 -.026 --- .000 .000 84.693 SDEV: .041 .061 .032 .042 .014 .033 .038 .020 .023 .204 .060 .005 --- .000 .000 .207 SERR: .012 .018 .009 .012 .004 .010 .011 .006 .007 .059 .017 .002 --- .000 .000 %RSD: 2.59 1.43 13.87-1520.63 33.07 17.23 .99 25.47 12.47 .31 .72 -21.27 --- .00 .00 STDS: 336 374 285 835 835 22 395 25 358 162 336 12 --- --- ---
Note that total is much better, but the SiO2 concentration is only 65.9 wt.%. Now we will recalculate the glass but this time include stoichiometric oxygen in the matrix correction:

Un 17 Withers-N5, Results in Oxide Weight Percents ELEM: Na2O K2O Cl BaO F TiO2 FeO MnO CaO SiO2 Al2O3 MgO O-D H2O O SUM 574 2.177 4.249 .205 .033 .085 .179 3.990 .077 .225 70.085 9.983 .016 --- .000 .000 91.303 575 2.144 4.244 .207 .018 .083 .206 3.974 .101 .205 69.801 9.909 .008 --- .000 .000 90.901 576 2.102 4.100 .180 .036 .099 .227 3.986 .107 .179 69.809 9.823 -.002 --- .000 .000 90.646 577 2.115 4.136 .258 .045 .109 .162 4.031 .074 .163 70.096 10.042 -.004 --- .000 .000 91.230 578 2.211 4.195 .258 -.015 .082 .119 4.041 .103 .161 69.906 9.904 .012 --- .000 .000 90.978 579 2.200 4.206 .239 -.087 .069 .206 4.064 .105 .181 69.881 10.022 .007 --- .000 .000 91.093 580 2.067 4.230 .223 .017 .087 .168 3.978 .070 .182 69.960 9.850 -.004 --- .000 .000 90.828 581 2.234 4.061 .248 -.021 .120 .195 3.949 .064 .211 70.160 9.897 .007 --- .000 .000 91.123 582 2.116 4.137 .164 -.028 .079 .222 4.076 .065 .166 70.273 9.939 .012 --- .000 .000 91.222 583 2.100 4.202 .218 -.068 .033 .244 4.032 .040 .198 70.295 9.953 .003 --- .000 .000 91.248 584 2.141 4.125 .213 .004 .033 .206 3.981 .082 .184 70.459 9.840 .003 --- .000 .000 91.271 585 2.076 4.141 .191 .032 .070 .200 4.010 .079 .223 70.423 9.943 .007 --- .000 .000 91.394 AVER: 2.140 4.169 .217 -.003 .079 .194 4.009 .080 .190 70.096 9.925 .005 --- .000 .000 91.103 SDEV: .055 .060 .030 .042 .026 .034 .039 .021 .022 .230 .069 .006 --- .000 .000 .223 SERR: .016 .017 .009 .012 .008 .010 .011 .006 .006 .066 .020 .002 --- .000 .000 %RSD: 2.55 1.44 13.88-1522.79 33.11 17.23 .98 25.47 11.70 .33 .70 120.12 --- .00 .00 STDS: 336 374 285 835 835 22 395 25 358 162 336 12 --- --- ---
Please note that by including the stoichiometric oxygen in the matrix correction the SiO2 concentration went from 65.9 wt.% to 70.09 wt.%. This is because oxygen absorbs Si Ka more than Si. So without stoichiometric oxygen in the matrix correction, the matrix is anomalously high in Si, which again, does not absorb Si Ka as much as oxygen, so the matrix correction for Si is underestimated.

But we still have a total of only 91.1 wt.%, so one might be forgiven to estimate the water by difference as 8.9 wt.%. But one would be wrong because we haven't provided a correction for the ion migration of Na and the effect on the other elements. The TDI correction for Na in this glass is around 100% as seen here in this plot of log intensity vs. time:

So turning on the TDI correction we obtain these analyses:

Un 17 Withers-N5, Results in Oxide Weight Percents ELEM: Na2O K2O Cl BaO F TiO2 FeO MnO CaO SiO2 Al2O3 MgO O-D H2O O SUM 574 4.117 4.182 .204 .033 .084 .179 3.989 .077 .224 69.392 10.084 .018 --- .000 .000 92.582 575 4.842 4.291 .207 .018 .082 .206 3.973 .101 .205 70.089 10.042 .010 --- .000 .000 94.065 576 4.255 3.997 .180 .036 .097 .227 3.985 .107 .179 69.839 9.930 -.001 --- .000 .000 92.831 577 4.575 4.261 .258 .045 .108 .162 4.029 .074 .163 69.847 10.167 -.002 --- .000 .000 93.688 578 4.550 4.211 .258 -.015 .081 .119 4.039 .103 .161 69.530 10.023 .013 --- .000 .000 93.073 579 4.687 4.256 .239 -.087 .068 .206 4.062 .105 .181 69.813 10.148 .008 --- .000 .000 93.685 580 4.257 4.207 .223 .017 .086 .168 3.976 .070 .182 69.621 9.960 -.003 --- .000 .000 92.765 581 4.968 4.191 .247 -.021 .118 .195 3.947 .064 .210 70.217 10.033 .008 --- .000 .000 94.177 582 4.447 4.237 .164 -.028 .079 .222 4.074 .065 .166 70.473 10.054 .013 --- .000 .000 93.966 583 4.547 4.070 .218 -.068 .032 .244 4.030 .040 .198 69.777 10.078 .004 --- .000 .000 93.170 584 4.524 4.049 .212 .004 .032 .206 3.979 .082 .184 70.249 9.959 .005 --- .000 .000 93.486 585 3.984 4.322 .191 .032 .069 .200 4.008 .078 .223 70.019 10.041 .008 --- .000 .000 93.176 AVER: 4.479 4.190 .217 -.003 .078 .194 4.008 .080 .190 69.905 10.043 .007 --- .000 .000 93.389 SDEV: .288 .101 .030 .042 .026 .033 .039 .020 .022 .317 .072 .006 --- .000 .000 .534 SERR: .083 .029 .009 .012 .007 .010 .011 .006 .006 .091 .021 .002 --- .000 .000 %RSD: 6.43 2.40 13.86-1523.55 33.08 17.23 .98 25.47 11.70 .45 .71 95.57 --- .00 .00 STDS: 336 374 285 835 835 22 395 25 358 162 336 12 --- --- ---
Note that the Na concentration went from around 2.2 wt% to 4.4 wt.%, but the Si dropped slightly because the TDI effect on Si is usually an increase in intensity as the Na intensity drops due to ion migration under the electron beam.

Now based on our totals of around 93.4 wt.%, our water by difference estimate would be around 6.6 wt.%, but that isn't very close to the FTIR H2O measurement of 5.06 wt.%. So what's going on? Well we have not yet included the water by difference into the matrix correction. Once we do that we obtain these results:

Un 17 Withers-N5, Results in Oxide Weight Percents ELEM: Na2O K2O Cl BaO F TiO2 FeO MnO CaO SiO2 Al2O3 MgO O-D H2O O SUM 574 4.191 4.211 .205 .033 .086 .180 4.037 .078 .227 70.047 10.230 .020 --- 6.454 .000 100.000 575 4.912 4.315 .207 .018 .084 .207 4.012 .102 .207 70.615 10.157 .011 --- 5.152 .000 100.000 576 4.329 4.024 .181 .036 .100 .229 4.032 .108 .181 70.477 10.069 .001 --- 6.233 .000 100.000 577 4.646 4.286 .259 .046 .110 .164 4.071 .075 .165 70.405 10.292 -.001 --- 5.483 .000 100.000 578 4.627 4.239 .259 -.015 .083 .120 4.085 .104 .163 70.140 10.158 .015 --- 6.022 .000 100.000 579 4.760 4.281 .240 -.088 .070 .207 4.104 .106 .182 70.371 10.272 .010 --- 5.485 .000 100.000 580 4.333 4.236 .224 .018 .088 .169 4.024 .070 .184 70.264 10.101 -.001 --- 6.290 .000 100.000 581 5.040 4.214 .248 -.021 .120 .196 3.985 .064 .212 70.735 10.146 .010 --- 5.052 .000 100.000 582 4.513 4.261 .164 -.028 .080 .224 4.114 .065 .168 71.014 10.173 .015 --- 5.237 .000 100.000 583 4.624 4.096 .218 -.069 .033 .246 4.075 .040 .200 70.382 10.213 .006 --- 5.936 .000 100.000 584 4.597 4.074 .213 .004 .033 .207 4.022 .083 .186 70.832 10.085 .007 --- 5.658 .000 100.000 585 4.050 4.350 .191 .033 .071 .202 4.053 .079 .225 70.631 10.176 .010 --- 5.929 .000 100.000 AVER: 4.552 4.216 .217 -.003 .080 .196 4.051 .081 .192 70.493 10.173 .009 --- 5.744 .000 100.000 SDEV: .288 .100 .030 .043 .026 .034 .039 .021 .022 .284 .070 .007 --- .469 .000 .000 SERR: .083 .029 .009 .012 .008 .010 .011 .006 .006 .082 .020 .002 --- .135 .000 %RSD: 6.33 2.38 13.86-1528.18 33.05 17.22 .97 25.48 11.66 .40 .69 76.35 --- 8.16 .00 STDS: 336 374 285 835 835 22 395 25 358 162 336 12 --- --- ---
Note that our SiO2 concentration went up slightly because we added additional oxygen (from H2O) into the matrix calculation, and so now our water by difference is around 5.7 wt.% which is quite close to our FTIR value of 5.06 wt.%. If we had included other trace elements in the analysis, we would get even closer.

So, the lesson is, always include all the unanalyzed elements into the matrix correction, or you will have significant errors in the analyzed elements. Excel is great, but it doesn't know physics! Please let me know if anyone has any questions.

pvburger · « **Reply #53 on:** March 08, 2019, 11:03:41 AM »

Does it make a difference if one includes oxygen as a specified element in the Elements/Cations menu of the Acquire! window prior to acquisition vs. simply selecting "Calculate with stoichiometric oxygen" under Calculation Options during the analysis step (following acquisition)?

I realize the output to the log may vary, but do the final concentrations saved from the Output menu differ? If so, is there any way to add "specified" oxygen as an element to a sample set that was initially acquired without it?

Cheers!

John Donovan · « **Reply #54 on:** March 08, 2019, 11:33:18 AM »

Quote from: pvburger on March 08, 2019, 11:03:41 AM

Does it make a difference if one includes oxygen as a specified element in the Elements/Cations menu of the Acquire! window prior to acquisition vs. simply selecting "Calculate with stoichiometric oxygen" under Calculation Options during the analysis step (following acquisition)?

I realize the output to the log may vary, but do the final concentrations saved from the Output menu differ? If so, is there any way to add "specified" oxygen as an element to a sample set that was initially acquired without it?

Cheers!

Hi Paul,
This is a really good question because it relates to an important point about how stoichiometric oxygen is handled along with "excess oxygen" in Probe for EPMA (and CalcZAF for that matter).

Basically, stoichiometric oxygen or oxygen by stoichiometry are unrelated to acquisition because in either case they are "unanalyzed" or "specified" elements. So you can specify them prior to acquisition or after acquisition of the "analyzed" elements.

If you check the "Use Stoichiometric Oxygen" option in the Calculation Options dialog, the software will automatically add oxygen to your matrix and include it in the matrix corrections.

But sometimes, we will want to specify some additional or "excess" oxygen as part of our analysis. For example your Fe is specified as FeO, but you are measuring a magnetite specimen. There is some discussion here on this:

https://probesoftware.com/smf/index.php?topic=223.msg1002#msg1002

But the bottom line is that yes, you can have oxygen calculated by stoichiometry, *and* you can add oxygen as a "specified" element for dealing with excess or oxygen deficits. Oxygen as an unanalyzed element is added, just as you would with any other unanalyzed elements, from the Elements/Cations dialog, but leaving the x-ray line blank.

Now all of this changes when you have oxygen as an analyzed element! When oxygen is being analyzed for (and when the oxidation state of Fe is uncertain this is actually a very good idea), the software will automatically subtract the stoichiometric oxygen from the measured oxygen to obtain the so called "excess" oxygen. This post describes that situation:

https://probesoftware.com/smf/index.php?topic=197.msg888#msg888

This can get complicated but play around with your samples, and if you can, try analyzing some standards as unknowns, e.g., magnetite, and it should become clear. But feel free to post some examples of what you are trying to accomplish if you have any questions at all.

Probeman · « **Reply #55 on:** July 02, 2019, 11:09:46 AM »

Most of us understand the effects of "unanalyzed" elements in the matrix correction and how it is important to include these elements either by specification, difference or by stoichiometry to other analyzed elements, in order to obtain an accurate matrix correction.

For example the issue of unanalyzed water in hydrous glasses, where one gets a results with a total of say 95% and one might be forgiven to think that, OK, I've got 5 wt% H2O there because 100 - 95 = 5 in Excel. But that would be wrong because Excel doesn't know anything about matrix correction physics. In fact what we see when we specify water by difference is described in this and other posts in this topic:

https://probesoftware.com/smf/index.php?topic=92.msg7701#msg7701

Basically adding oxygen (and hydrogen) to our glass matrix correction results in a significant change in the concentration of other elements. In a generic example of a silicate glass, the Si (and other element) concentrations increases significantly. Why is this? Because it turns out that Si Ka is not well absorbed by Si atoms, but they are well absorbed by oxygen atoms! Once the matrix correction "knows" about the extra oxygen (water), the absorption correction increases correspondingly.

So in our generic example of a 95% total in a hydrous glass, once we add H2O by difference and include it in the matrix correction, because the Si (and other elements) increase by about 1%, we instead obtain about 4% H2O by difference. This effect was described in the Roman et al. paper referenced here:

https://probesoftware.com/smf/index.php?topic=61.msg4303#msg4303

Now this makes some intuitive sense to me because Si Ka is a fairly low energy emission line and therefore fairly well absorbed by other elements, but I never expected this to be the case for Fe Ka at 6.4 keV!

Here's an example for a magnetite sample where the Fe is expressed, as we usually do in EPMA, as FeO:

Un 96 7138_PPO-2_Mgt4_HO-1, 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 EXP LIN MAN MAN TIME: 40.00 120.00 40.00 80.00 30.00 30.00 80.00 120.00 --- BEAM: 45.93 45.93 45.93 45.93 45.93 45.93 45.93 45.93 --- ELEM: Fe Mg Si Ti V Mn Cr Al O SUM 146 60.944 1.463 .058 5.329 .265 .422 .375 1.542 23.842 94.240 AVER: 60.944 1.463 .058 5.329 .265 .422 .375 1.542 23.842 94.240 SDEV: .000 .000 .000 .000 .000 .000 .000 .000 .000 .000 SERR: .000 .000 .000 .000 .000 .000 .000 .000 .000 %RSD: .00 .00 .00 .00 .00 .00 .00 .00 .00 STDS: 396 396 14 22 23 25 396 396 --- STKF: .1836 .0330 .4101 .5547 .6328 .7341 .3060 .0469 --- STCT: 1119.3 3499.0 4307.2 9290.4 8344.6 14435.6 1338.8 1992.0 --- UNKF: .5710 .0066 .0004 .0548 .0028 .0040 .0043 .0088 --- UNCT: 3481.1 701.1 4.3 917.6 37.1 77.9 18.9 373.9 --- UNBG: 12.2 47.8 2.8 32.9 15.3 29.2 6.1 29.4 --- ZCOR: 1.0674 2.2108 1.4127 .9728 .9416 1.0658 .8682 1.7520 --- KRAW: 3.1102 .2004 .0010 .0988 .0044 .0054 .0141 .1877 --- PKBG: 286.31 15.65 2.55 28.88 3.43 3.67 4.10 13.70 --- INT%: .00 ---- ---- ---- -15.76 ---- -1.38 ---- ---
As one can see our Fe concentration is 60.944 and our total is 94.240. So we're missing about 5% of something and that something of course is mostly the excess oxygen in the Fe2O3 molecule in magnetite. Now how much of an effect can this ~5% missing oxygen have on our Fe concentration? What would you guess? I wouldn't have guessed this much:

Un 96 7138_PPO-2_Mgt4_HO-1, Results in Elemental Weight Percents ELEM: Fe Mg Si Ti V Mn Cr Al O TYPE: ANAL ANAL ANAL ANAL ANAL ANAL ANAL ANAL DIFF BGDS: MAN MAN MAN MAN EXP LIN MAN MAN TIME: 40.00 120.00 40.00 80.00 30.00 30.00 80.00 120.00 --- BEAM: 45.93 45.93 45.93 45.93 45.93 45.93 45.93 45.93 --- ELEM: Fe Mg Si Ti V Mn Cr Al O SUM 146 61.421 1.451 .057 5.379 .267 .425 .383 1.532 29.085 100.000 AVER: 61.421 1.451 .057 5.379 .267 .425 .383 1.532 29.085 100.000 SDEV: .000 .000 .000 .000 .000 .000 .000 .000 .000 .000 SERR: .000 .000 .000 .000 .000 .000 .000 .000 .000 %RSD: .00 .00 .00 .00 .00 .00 .00 .00 .00 STDS: 396 396 14 22 23 25 396 396 --- STKF: .1836 .0330 .4101 .5547 .6328 .7341 .3060 .0469 --- STCT: 1119.3 3499.0 4307.2 9290.4 8344.6 14435.6 1338.8 1992.0 --- UNKF: .5710 .0066 .0004 .0548 .0028 .0040 .0044 .0088 --- UNCT: 3481.6 701.6 4.3 918.5 37.1 77.9 19.1 374.1 --- UNBG: 11.8 47.3 2.8 32.0 15.3 29.2 5.9 29.2 --- ZCOR: 1.0756 2.1904 1.4061 .9808 .9506 1.0740 .8773 1.7396 --- KRAW: 3.1106 .2005 .0010 .0989 .0044 .0054 .0143 .1878 --- PKBG: 296.98 15.83 2.54 29.68 3.43 3.67 4.25 13.80 --- INT%: .00 ---- ---- ---- -15.75 ---- -1.36 ---- ---
Holy cow! Our Fe concentration went from 60.944 to 61.421 which is almost 0.5 wt% absolute. And note how the ZCOR (matrix correction) for Fe Ka went from 1.0674 to 1.0756. But even more mind blowing is to compare the *other* elements we measured: Mg, Ti, Al, etc. What happened to them? Well they went slightly *down* in concentration! Why? Because they are more absorbed by Fe than by O. It just goes to show you that one cannot intuit physics, you have to just run the darn calculation.

Of course we really should calculate the excess oxygen in our magnetites/ilmenites using a ferric/ferrous calculation (and more on that later), but the oxygen by difference calculation demonstrates that once we obtain our ferric/ferrous ratio, we really need to calculate the excess oxygen from that and specify it in Probe for EPMA, to obtain a more accurate Fe concentration for our mineral thermodynamic calculations.

Now, I'm no geologist but perhaps one should then recalculate our ferric/ferrous ratios, this time using our improved Fe concentration. Again, more on this later...

Probeman · « **Reply #56 on:** July 11, 2019, 12:39:23 PM »

I'm no geologist but I think this is a cool trick to deal with missing OH in (some/all?) amphiboles.

So here is a tremolite (asbestos) analysis I did recently for the CDC:

Un 13 AZ asbestos gr6, Results in Oxide Weight Percents ELEM: Na2O SiO2 K2O Al2O3 FeO MgO CaO Cl TiO2 F MnO ZnO O HO SUM 642 -.016 57.171 .112 .186 3.613 22.006 12.782 .004 -.044 -.063 .185 .028 .000 .000 95.963 643 .006 57.423 .086 .203 3.307 22.143 12.787 .000 -.031 .049 .215 .009 .000 .000 96.199 644 .008 57.522 .078 .182 3.437 22.184 12.780 .004 -.015 -.027 .178 .026 .000 .000 96.358 645 .060 57.454 .073 .163 3.168 22.240 12.908 .003 .012 -.071 .175 .029 .000 .000 96.213 646 .088 57.484 .079 .053 3.012 22.158 13.087 -.003 -.036 -.047 .138 -.041 .000 .000 95.972 AVER: .029 57.411 .086 .157 3.307 22.146 12.869 .002 -.023 -.032 .178 .010 .000 .000 96.141 SDEV: .043 .139 .016 .060 .233 .086 .134 .003 .022 .049 .028 .030 .000 .000 .170 SERR: .019 .062 .007 .027 .104 .039 .060 .001 .010 .022 .012 .013 .000 .000 %RSD: 146.89 .24 18.18 38.14 7.04 .39 1.04 159.31 -97.92 -151.68 15.48 295.57 223.61 .00 STDS: 336 162 374 336 162 162 162 285 22 835 25 660 --- ---
Note that the total is quite low due to the missing OH molecule. Now looking up the formula for tremolite we see this: Ca2Mg5Si8O22(OH)2. So there are 2 OH molecules for every 8 Si atoms or 0.25 hydroxyls for every Si atom.

Next we add hydrogen as an unanalyzed elements in the Elements/Cations dialog like this, being sure to specify 1 hydrogen and 1 oxygen, to obtain the hydroxyl molecule:

Next we go into the Calculation Options dialog and specify oxygen by stoichiometry, and .25 hydrogens (as hydroxyl), to 1 Si atom as seen here:

Now when we calculate the composition we get:

Un 13 AZ asbestos gr6, Results in Oxide Weight Percents ELEM: Na2O SiO2 K2O Al2O3 FeO MgO CaO Cl TiO2 F MnO ZnO O HO SUM 642 -.014 57.329 .112 .187 3.634 22.208 12.820 .004 -.044 -.065 .186 .028 .000 4.057 100.442 643 .009 57.587 .086 .204 3.326 22.349 12.826 .000 -.032 .050 .216 .009 .000 4.075 100.705 644 .011 57.682 .078 .183 3.456 22.389 12.819 .004 -.015 -.028 .179 .026 .000 4.082 100.866 645 .063 57.613 .073 .164 3.186 22.446 12.946 .003 .012 -.073 .176 .029 .000 4.077 100.717 646 .091 57.650 .079 .054 3.030 22.366 13.126 -.003 -.036 -.048 .139 -.042 .000 4.080 100.486 AVER: .032 57.572 .086 .158 3.326 22.352 12.908 .002 -.023 -.033 .179 .010 .000 4.074 100.643 SDEV: .043 .141 .016 .060 .234 .088 .134 .003 .022 .049 .028 .030 .000 .010 .176 SERR: .019 .063 .007 .027 .105 .039 .060 .001 .010 .022 .012 .013 .000 .004 %RSD: 136.72 .24 18.19 38.03 7.03 .39 1.04 159.31 -97.92 -151.40 15.48 295.57 -651.92 .24 STDS: 336 162 374 336 162 162 162 285 22 835 25 660 --- ---
Not only is our total now very close to 100%, but note how the Si and Mg compositions have changed, because now the matrix effect of that 4 wt% OH was included in the matrix calculations.

Brian Joy · « **Reply #57 on:** July 11, 2019, 01:21:49 PM »

Quote from: Probeman on July 11, 2019, 12:39:23 PM

I'm no geologist but I think this is a cool trick to deal with missing OH in (some/all?) amphiboles.

So here is a tremolite (asbestos) analysis I did recently for the CDC:

Un 13 AZ asbestos gr6, Results in Oxide Weight Percents ELEM: Na2O SiO2 K2O Al2O3 FeO MgO CaO Cl TiO2 F MnO ZnO O HO SUM 642 -.016 57.171 .112 .186 3.613 22.006 12.782 .004 -.044 -.063 .185 .028 .000 .000 95.963 643 .006 57.423 .086 .203 3.307 22.143 12.787 .000 -.031 .049 .215 .009 .000 .000 96.199 644 .008 57.522 .078 .182 3.437 22.184 12.780 .004 -.015 -.027 .178 .026 .000 .000 96.358 645 .060 57.454 .073 .163 3.168 22.240 12.908 .003 .012 -.071 .175 .029 .000 .000 96.213 646 .088 57.484 .079 .053 3.012 22.158 13.087 -.003 -.036 -.047 .138 -.041 .000 .000 95.972 AVER: .029 57.411 .086 .157 3.307 22.146 12.869 .002 -.023 -.032 .178 .010 .000 .000 96.141 SDEV: .043 .139 .016 .060 .233 .086 .134 .003 .022 .049 .028 .030 .000 .000 .170 SERR: .019 .062 .007 .027 .104 .039 .060 .001 .010 .022 .012 .013 .000 .000 %RSD: 146.89 .24 18.18 38.14 7.04 .39 1.04 159.31 -97.92 -151.68 15.48 295.57 223.61 .00 STDS: 336 162 374 336 162 162 162 285 22 835 25 660 --- ---
Note that the total is quite low due to the missing OH molecule. Now looking up the formula for tremolite we see this: Ca2Mg5Si8O22(OH)2. So there are 2 OH molecules for every 8 Si atoms or 0.25 hydroxyls for every Si atom.

Next we add hydrogen as an unanalyzed elements in the Elements/Cations dialog like this, being sure to specify 1 hydrogen and 1 oxygen, to obtain the hydroxyl molecule:

Next we go into the Calculation Options dialog and specify oxygen by stoichiometry, and .25 hydrogens (as hydroxyl), to 1 Si atom as seen here:

Now when we calculate the composition we get:

Un 13 AZ asbestos gr6, Results in Oxide Weight Percents ELEM: Na2O SiO2 K2O Al2O3 FeO MgO CaO Cl TiO2 F MnO ZnO O HO SUM 642 -.014 57.329 .112 .187 3.634 22.208 12.820 .004 -.044 -.065 .186 .028 .000 4.057 100.442 643 .009 57.587 .086 .204 3.326 22.349 12.826 .000 -.032 .050 .216 .009 .000 4.075 100.705 644 .011 57.682 .078 .183 3.456 22.389 12.819 .004 -.015 -.028 .179 .026 .000 4.082 100.866 645 .063 57.613 .073 .164 3.186 22.446 12.946 .003 .012 -.073 .176 .029 .000 4.077 100.717 646 .091 57.650 .079 .054 3.030 22.366 13.126 -.003 -.036 -.048 .139 -.042 .000 4.080 100.486 AVER: .032 57.572 .086 .158 3.326 22.352 12.908 .002 -.023 -.033 .179 .010 .000 4.074 100.643 SDEV: .043 .141 .016 .060 .234 .088 .134 .003 .022 .049 .028 .030 .000 .010 .176 SERR: .019 .063 .007 .027 .105 .039 .060 .001 .010 .022 .012 .013 .000 .004 %RSD: 136.72 .24 18.19 38.03 7.03 .39 1.04 159.31 -97.92 -151.40 15.48 295.57 -651.92 .24 STDS: 336 162 374 336 162 162 162 285 22 835 25 660 --- ---
Not only is our total now very close to 100%, but note how the Si and Mg compositions have changed, because now the matrix effect of that 4 wt% OH was included in the matrix calculations.

Hi John,

This is not correct. Amphiboles generally have around 2 wt% H2O assuming that no F-, Cl-, or O2- substitutes for OH-. For charge balance, you should have specified 2 hydrogens per oxygen. Also, note that Al3+ can substitute for Si4+ on the tetrahedral sites, and so it is not sound practice to specify a given number of OH- per Si4+.

Brian

AndrewLocock · « **Reply #58 on:** July 11, 2019, 01:25:56 PM »

Hello,
The general idea of calculating the hydroxyl content in advance of the matrix corrections is a good one, but I should draw your attention to some problems with the discussion immediately above.

1) The concentrations of the elements should be expressed as neutral oxides.
Thus, the method should report percent-by-weight of H2O, not the hydroxyl moiety.
For both of the average tremolite compositions discussed, the amount of H2O(!) should be 2.19 wt%.
This corresponds to two hydroxyl units per formula unit in each of these cases:
(K0.016) (Ca1.923Mn0.021Fe0.019Na0.008)Σ1.971 (Mg4.605Fe0.367Al0.026Zn0.002)Σ5 (Si8.006)Σ8.006 O22 (OH)2
(K0.016) (Ca1.921Fe0.037Mn0.021Na0.008)Σ1.987 (Mg4.627Fe0.349Al0.021Zn0.002)Σ4.999 (Si7.995Al0.005)Σ8 O22 (OH)2
The final totals are thus 98.4 to 98.8 wt%.

2) Many (most) amphibole species do not have exactly 8 Si per formula unit.
One could instead use the ratio of H to oxygen for those amphiboles that are assumed to have only hydroxyl present (no oxo-substitution, and no significant F or Cl present).
The ratio would be 2 H for 24 oxygen, or 0.0833333 H for 1 O.

3) A minor point, but in my view, negative values of the measured elements should be automatically set to zero, prior to averaging.

Best regards,
Andrew

Probeman · « **Reply #59 on:** July 11, 2019, 01:55:31 PM »

As I said, I'm no geologist, so please explain why the formula for tremolite is given as Ca2Mg5Si8O22(OH)2

http://www.webmineral.com/data/Tremolite.shtml#.XSeiNP57laQ

I only assumed Si might be a valid element to ratio hydrogen to because there's almost no Al in this composition. Maybe this only works for tremolite?

My main point was just to show the matrix effect of including OH or H2O, on the other measured elements. What is your suggestion for adding in the missing hydroxyl or water to improve the matrix correction accuracy? Just specify 2% H2O?

News:

Author Topic: Specifying Unanalyzed Elements For a Proper Matrix Correction (Read 47977 times)

John Donovan

Re: Specifying Unanalyzed Elements For a Proper Matrix Correction

John Donovan

Re: Specifying Unanalyzed Elements For a Proper Matrix Correction

AndrewLocock

Re: Specifying Unanalyzed Elements For a Proper Matrix Correction

John Donovan

Re: Specifying Unanalyzed Elements For a Proper Matrix Correction

AndrewLocock

Re: Specifying Unanalyzed Elements For a Proper Matrix Correction

John Donovan

Re: Specifying Unanalyzed Elements For a Proper Matrix Correction

Probeman

Re: Specifying Unanalyzed Elements For a Proper Matrix Correction

Probeman

Re: Specifying Unanalyzed Elements For a Proper Matrix Correction

pvburger

Re: Specifying Unanalyzed Elements For a Proper Matrix Correction

John Donovan

Re: Specifying Unanalyzed Elements For a Proper Matrix Correction

Probeman

Re: Specifying Unanalyzed Elements For a Proper Matrix Correction

Probeman

Re: Specifying Unanalyzed Elements For a Proper Matrix Correction

Brian Joy

Re: Specifying Unanalyzed Elements For a Proper Matrix Correction

AndrewLocock

Re: Specifying Unanalyzed Elements For a Proper Matrix Correction

Probeman

Re: Specifying Unanalyzed Elements For a Proper Matrix Correction