But if the primary standard will have coating with thickness (chosen in PFE) 199 and less or 201 and more A, it calculated composition will change. Why the calculated composition will not change with thickness (chosen in PFE) exactly 200A?
Hi Rom,
You are exactly correct!
Apparently I've only tested situations where the standard was uncoated and the unknown was coated with various coatings, or both the standard and the unknown were coated the same (various) coatings, or the standards were coated with the default value of 200 angstroms of carbon and the unknowns coated with (various) coatings (generally the situation in my lab!), and of course for all of these scenarios the coating corrections were turned on.
In the situation you described where the primary standard was coated with a value different than the default coating (200 angstroms of carbon), *and* the standard was analyzed as an unknown *and* the coating correction flags in the Analytical | Analysis Options dialog were turned on, then a problem can be seen. Though it's possible that the difference in the coatings thicknesses were small enough to not be noticed, so "good catch"!
Of course normally we use the same coating on both standards and unknowns and usually we never turn on the coating corrections, so all this time no one (until you) has noticed this issue.
Update Probe for EPMA from the Help menu and you will get this coating bug fix for standards analyzed as unknowns with non-default coatings specified.
Basically, the problem was in the calculation of the "primary" (pure element intensities in routine ZAFSetZAF) when the sample was a standard with a non-standard coating thickness and analyzed as an unknown.
Here is what the old code showed when the standard was coated with 200 angstroms of carbon and analyzed with the coating corrections turned on:
Using Conductive Coating Correction For Electron Absorption and X-Ray Transmission:
Sample Coating=C, Density=1.35 gm/cm3, Thickness=200 angstroms, Sin(Thickness)=311.145 angstroms
St 522 Set 2 Titanium metal, Results in Elemental Weight Percents
ELEM: Ti O
TYPE: ANAL ANAL
BGDS: LIN LIN
TIME: 10.00 10.00
BEAM: 30.01 30.01
ELEM: Ti O SUM
106 99.276 .711 99.988
107 99.842 .679 100.521
108 99.300 .609 99.908
109 99.230 .554 99.784
110 99.987 .639 100.626
AVER: 99.527 .638 100.165
SDEV: .359 .061 .381
SERR: .160 .027
%RSD: .36 9.60
PUBL: 99.500 .500 100.000
%VAR: (.03) 27.63
DIFF: (.03) .138
STDS: 522 12
No problems with this analysis. And here with the standard coating set to 190 angstroms (again with the old code):
Using Conductive Coating Correction For Electron Absorption and X-Ray Transmission:
Sample Coating=C, Density=1.35 gm/cm3, Thickness=190 angstroms, Sin(Thickness)=295.5877 angstroms
St 522 Set 2 Titanium metal, Results in Elemental Weight Percents
ELEM: Ti O
TYPE: ANAL ANAL
BGDS: LIN LIN
TIME: 10.00 10.00
BEAM: 30.01 30.01
ELEM: Ti O SUM
106 99.333 .713 100.047
107 99.900 .681 100.580
108 99.357 .611 99.967
109 99.287 .555 99.842
110 100.045 .641 100.685
AVER: 99.584 .640 100.224
SDEV: .359 .061 .382
SERR: .160 .027
%RSD: .36 9.60
PUBL: 99.500 .500 100.000
%VAR: (.08) 28.02
DIFF: (.08) .140
STDS: 522 12
OK, it is a little different, but perhaps we can see why this was never noticed previously!
Now the same analysis but with the corrected code where the actual standard coating parameters are now loaded in the "primary intensity" calculation when the sample is a standard:
Using Conductive Coating Correction For Electron Absorption and X-Ray Transmission:
Sample Coating=C, Density=1.35 gm/cm3, Thickness=190 angstroms, Sin(Thickness)=295.5877 angstroms
St 522 Set 2 Titanium metal, Results in Elemental Weight Percents
ELEM: Ti O
TYPE: ANAL ANAL
BGDS: LIN LIN
TIME: 10.00 10.00
BEAM: 30.01 30.01
ELEM: Ti O SUM
106 99.276 .711 99.988
107 99.842 .679 100.521
108 99.300 .609 99.908
109 99.230 .554 99.784
110 99.987 .639 100.626
AVER: 99.527 .638 100.165
SDEV: .359 .061 .381
SERR: .160 .027
%RSD: .36 9.60
PUBL: 99.500 .500 100.000
%VAR: (.03) 27.63
DIFF: (.03) .138
STDS: 522 12
Now we get the same result as before when all standards are coated with 200 angstroms of carbon! And finally here again, but
without any coating corrections turned on:
No Sample Coating and/or No Sample Coating Correction
St 522 Set 2 Titanium metal, Results in Elemental Weight Percents
ELEM: Ti O
TYPE: ANAL ANAL
BGDS: LIN LIN
TIME: 10.00 10.00
BEAM: 30.01 30.01
ELEM: Ti O SUM
106 99.276 .711 99.988
107 99.842 .679 100.521
108 99.300 .609 99.908
109 99.230 .554 99.784
110 99.987 .639 100.626
AVER: 99.527 .638 100.165
SDEV: .359 .061 .381
SERR: .160 .027
%RSD: .36 9.60
PUBL: 99.500 .500 100.000
%VAR: (.03) 27.63
DIFF: (.03) .138
STDS: 522 12
Again, the same result because all standards have the same coating!
Remember, in situations of very high x-ray absorption by the coating, or in very low over voltage situations, the elemental compositions may be somewhat different due to differences in the coating physics between the primary and secondary (or unknown) compositions.