Standard/ unknown different coating?

[pgopon]

Hi all,

I have been playing with the feature in PfEPMA where you can designate different coatings for standards/unknowns and was wondering if anyone knew how well or if this feature worked.

I have attached a excel file that shows the results of the PfEPMA matrix correction with and without the coating correction.  In this case my carbon standard was un-coated, the rest of the standards were C coated, with the exception of the Fe and Si standards were Pt coated at the same time as the unknown.  I would have thought add/removing this correction would have a larger effect on the wt %, does anyone have any insight as to why the correction is so small.  I have done some of the monte carlo modeling and we should be seeing more counts on the platinum coated sample.

Cheers,

phil

Edit by JJD to add Excel sheet as attachment

[John Donovan]

The coating correction in PFE is quite a crude correction.  One part is an absorption correction for low energy x-rays emitted through the coating and the second, is a correction for electron energy loss from primary electrons going through the coating.

It does not account for additional emitted x-rays from the coating.

[John Donovan]

An improved coating correction has been implemented in Probe for EPMA (and CalcZAF) based on a modified coating correction by Kerrick, et. al., Amer. Min. 58, 920-925 (1973). You'll need v. 10.3.6 to utilize these new methods.

There are two parts to this coating correction:

1. A correction for "electron absorption" which compensates for the decrease in over voltage as the beam decelerates through the coating thickness.

' Intensity correction from electron energy loss due to coating
ratio! = (100# - ((8.3 * sample(1).CoatingDensity! * sample(1).CoatingThickness! / ANGPERNM&) / (sample(1).KilovoltsArray!(chan%) ^ 2 - (sample(1).LineEdge!(chan%) / EVPERKEV#) ^ 2))) / 100#

2. A correction for x-ray attenuation as the emitted x-ray transmits through the coating layer (coating thickness / Sin (theta)).

' Calculate thickness based on takeoff angle
radians! = sample(1).takeoff! * PI! / 180#
thickness! = sample(1).CoatingThickness! / Sin(radians!)

' Calculate x-ray transmission based on Sin thickness in angstroms
transmission! = NATURALE# ^ (-1# * atotal! * sample(1).CoatingDensity! * thickness! * CMPERANGSTROM#)


A test sample was prepared of an ultra pure Ti metal coated with pure Ag. The thickness was determined by XRR as 120 nm as shown in the following diffraction fringes:



A separate determination was performed using StrataGEM using 6, 8 and 15 keV which gave 124.7 nm assuming a density of 10.5 for Ag and no carbon coat on the pure Ti or the Ag coated Ti.





Results from initial EPMA measurements produce the following k-ratios:

6 keV       0.0298
8 keV       0.0795
15 keV     0.463

Results from MC modeling and PFE coating corrections will be posted later this week and thanks to John Armstrong for his assistance.

[John Donovan]

The coating correction is specified for both standards and unknowns within each probe run database. This way, your coating specifications cannot be modified unless you want to. For example to specify the coating parameters for your standards separately or all at once using the Standard | Edit Standard Parameters (coating) menu as seen here:



For your unknowns the coating parameters can also be assign to each unknown or all selected as seen here:



These flags are always defaulted to the coating parameters defined in the Probewin.ini file, but can be modified by editing these keywords:

[standards]
StandardCoatingFlag=1       ; 0 = not coated, 1 = coated
StandardCoatingElement=6     ; assume carbon
StandardCoatingDensity=2.1
StandardCoatingThickness=200   ; in angstroms

and the coating flags for electron absorption and x-ray transmission can be turned on from the Analytical | Analysis Options as seen here:

[Probeman]

I re-ran the Ag coated Ti sample at 8, 9, 10, 11, 12, 13, 14, 15, 20 and 25 keV on my SX50 and after processing with STRATAGem* I obtained 119.6 nm. This compares quite well to the XRR measurement of 119.7 nm!

I call this "spurious accuracy" (credit to Joe Michael for the concept)!  That is, too good to be true!   

In fact the STRATAGem* keV vs. K-ratio plot seems to show a small but systematic bias in the data. The student that performed the deposition and XRR thinks this is due to the Ag density at the surface of the Ti not being exactly 10.5.  But for the purposes of a coating correction let us assume that the Ag film is 120 nm.

See attached documents below.

* © Copyright 1993-2016 SAMx

[Probeman]

Here are some preliminary results comparing 6, 8 and 15 keV measurements of Ti Ka on a Ti sample coated with 120 nm of Ag (see attached spreadsheet).



As you can see the MC calculations with an assumed coating thickness of 120 nm are close, but the MC k-ratios seem to indicate a coating thickness of a little over 100 nm thickness for the Ag.

Be that as it may, we can now start to test the coating correction in Probe for EPMA... but remember, this coating correction was calibrated for 20 nm of carbon, not 120 nm of Ag!  So we can assume that it will fail, but how bad will it fail and can this coating correction be improved further?

[Probeman]

Let's start with what should be the easiest coating correction for this sample, Ti Ka underneath 120 nm of Ag, at 25 keV. Without the coating correction, we obtain the following:

Un   68 coated Ti (MVA)
TakeOff = 40.0  KiloVolt = 25.0  Beam Current = 30.0  Beam Size =    0
(Magnification (analytical) =  20000),        Beam Mode = Analog  Spot
(Magnification (default) =      400, Magnification (imaging) =    400)
Image Shift (X,Y):                                          .00,   .00
Pre Acquire String :                                        mag 100000
Post Acquire String :                                          mag 400
Number of Data Lines:   5             Number of 'Good' Data Lines:   5
First/Last Date-Time: 05/22/2014 03:19:44 AM to 05/22/2014 03:32:22 AM

Average Total Oxygen:         .000     Average Total Weight%:   68.318
Average Calculated Oxygen:    .000     Average Atomic Number:   22.000
Average Excess Oxygen:        .000     Average Atomic Weight:   47.900
Average ZAF Iteration:        1.00     Average Quant Iterate:     2.00

Un   68 coated Ti (MVA), Results in Elemental Weight Percents
 
ELEM:       Ti
BGDS:      LIN
TIME:    80.00
BEAM:    30.14

ELEM:       Ti   SUM 
   305  68.423  68.423
   306  68.206  68.206
   307  68.639  68.639
   308  68.538  68.538
   309  67.784  67.784

AVER:   68.318  68.318
SDEV:     .339    .339
SERR:     .152
%RSD:      .50
STDS:     2522


With a 68% total (relative to pure Ti) we can see that 120 nm of Ag significantly affects the production and emission of Ti Ka x-rays. Now we'll turn the correction on as discussed here above:

http://probesoftware.com/smf/index.php?topic=23.msg1258#msg1258

and now we obtain the following results corrected for the 120 nm of Ag on Ti:

Un   68 coated Ti (MVA)
TakeOff = 40.0  KiloVolt = 25.0  Beam Current = 30.0  Beam Size =    0
(Magnification (analytical) =  20000),        Beam Mode = Analog  Spot
(Magnification (default) =      400, Magnification (imaging) =    400)
Image Shift (X,Y):                                          .00,   .00
Pre Acquire String :                                        mag 100000
Post Acquire String :                                          mag 400
Number of Data Lines:   5             Number of 'Good' Data Lines:   5
First/Last Date-Time: 05/22/2014 03:19:44 AM to 05/22/2014 03:32:22 AM

Average Total Oxygen:         .000     Average Total Weight%:   99.101
Average Calculated Oxygen:    .000     Average Atomic Number:   22.000
Average Excess Oxygen:        .000     Average Atomic Weight:   47.900
Average ZAF Iteration:        1.00     Average Quant Iterate:     2.00

Using Conductive Coating Correction For Electron Absorption and X-Ray Transmission:
Sample Coating=Ag, Density=10.5 gm/cm3, Thickness=1200 angstroms, Sin(Thickness)=1866.87 angstroms

Un   68 coated Ti (MVA), Results in Elemental Weight Percents
 
ELEM:       Ti
BGDS:      LIN
TIME:    80.00
BEAM:    30.14

ELEM:       Ti   SUM 
   305  99.253  99.253
   306  98.938  98.938
   307  99.567  99.567
   308  99.420  99.420
   309  98.326  98.326

AVER:   99.101  99.101
SDEV:     .492    .492
SERR:     .220
%RSD:      .50
STDS:     2522


Not bad at all, in fact it looks like "spurious accuracy" too good to be true, but let's look at the other lower voltages here in this table, which we expect to be more significant coating corrections:



Clearly we got lucky on the 25 keV correction, but the lower voltages increasingly produce an over correction in the Ti concentrations, at least until the electron absorption correction goes negative!

Clearly this 20 nm carbon coating correction can be improved. Suggestions...?

[Gseward]

John,

I am a little confused about when the modifications to the coating parameters are applied. I open the Standards coating parameters dialog from the Standard menu, modify my standard coatings (some coated with 20nm C, some uncoated), close the dialog and then recalculate the analysis, but the modified coating parameters don't seem to have been applied? What am I doing wrong?

Having modified the coatings, if I select all samples (unkn and Sts) in Analyze! open the calculations options window, then close it again, the modifications are applied, but this seems like a round about way of updating the changes.

Please enlighten me!

gareth

[John Donovan]

Happy to do so!

Basically PFE loads defaults for the standard and unknown coating parameters based on the values in the Probewin.ini file.

Then those options are disabled at a global level (you'll see why in a minute), so they are loaded but not utilized by default in the matrix corrections.

To turn on the coating parameters explicitly you need to go to the Analytical | Analysis Option menu dialog and check these options:



Because the user will generally only be modifying their unknown coating parameters, this global flag is turned on automatically if the user edits the coating parameters in the Calculation Options dialog for unknown samples. This is explained in the post here above:

http://probesoftware.com/smf/index.php?topic=23.msg1258#msg1258

This action is similar to how specifying a spectral interference correction in the Standard Assignments dialog also turns on the global interference correction flag in Analysis Options.

Why do it like this?  So every correction the software performs can be toggled on/off without having to re-specify all the sample level parameters!

Question for you:  does your version have the new purple interactive Help button visible?  That is in the latest version and takes you right here in the forum!

[Gseward]

Thanks for the reply, John, but this doesn't really answer my question. When I modify the standard coatings, the new parameters are not applied, unless I go through the route I described previously. 

I am running the current version, purple buttons and all!

I don't have to modify standard coatings very often, so I'll not worry about it any further.

Gareth

[John Donovan]

Thanks for the reply, John, but this doesn't really answer my question. When I modify the standard coatings, the new parameters are not applied, unless I go through the route I described previously. 

Right, exactly. That is a feature (as opposed a bug)- because one doesn't often modify the standard coatings!

In other words- I want the user to specifically want the coating correction applied when simply browsing the standard coating parameters.
john

[JohnF]

JohnD:

Can you please clarify: if the two boxes for Coating Corrections (electron energy,  X-ray absorption) are NOT checked in Analysis Calculations Options, while meanwhile there are default settings for Carbon coating on both std and unknowns, is NO correction done for the carbon coating? And ONLY when the two boxes are checked, and only then, is there a correction (albeit small) made?

thanks.

JohnF

[Probeman]

JohnD:

Can you please clarify: if the two boxes for Coating Corrections (electron energy,  X-ray absorption) are NOT checked in Analysis Calculations Options, while meanwhile there are default settings for Carbon coating on both std and unknowns, is NO correction done for the carbon coating? And ONLY when the two boxes are checked, and only then, is there a correction (albeit small) made?

thanks.

JohnF

JohnF,
That is correct.

The coating element, thickness, on/off flags in the Analyze! | Calculation Options dialog are specified on a sample basis.  The Analytical | Analysis Options menu dialog are a way to turn all these sample coating calculations on/off on a global basis.
JohnD

[Probeman]

I should add that you will notice that the coating options in the Analyze! | Calculation Options dialog are disabled for standard samples. 

This is because coating parameters for standards are specified from the Standard | Edit Standard Coating Parameters menu dialog as seen here:



john

[UofO EPMA Lab]

Probeman here.

While at M&M last week I was asked a question about analyzing standards in Probe for EPMA. That is one acquires data on a standard and then from the Analyze! window one clicks the Analyze button to get a quantitative analysis of that standard *as though it were an unknown*.

Of course if the standard is the primary (assigned) standard for that element, the element in question will iterate to the correct concentration, because the standard is analyzing itself as the standard. That is why the %VAR (percent variance) value for that element is displayed in parentheses- because it isn't really a test of the analytical quality. 

More interesting is when one analyzes a standard (as an unknown from the Analyze! window), and that standard contains a non-zero concentration of an element (or a zero concentration to check for spectral interferences!), and that standard is *not* the primary standard for the element in question. Then we have the so-called secondary standard situation which is the best way to check for accuracy in EPMA.

And of course because the program knows that this sample is a standard, it can also print out the published analysis of the standard from the standard database for comparison.

But it is important to remember that for the quantification of unknowns, the only elements utilized in the primary standard are the elements that are actually assigned as the primary standard for that element.  All the other elements in that standard are only useful for accuracy evaluation of major elements, or zero concentrations for interference corrections, or of course as standards for the MAN background correction.

So, when you have the situation where the standards are coated with, for example carbon, and the unknowns are not coated, say for carbon analysis of a steel, you will be analyzing the coating. So, the standard for iron might be pure Fe, and one would get an analysis (from the Analyze! window) like this:

St  526 Set   3 Iron metal

St  526 Set   3 Iron metal
TakeOff = 40.0  KiloVolt = 15.0  Beam Current = 50.0  Beam Size =    0
(Magnification (analytical) =  40000),        Beam Mode = Analog  Spot
(Magnification (default) =      400, Magnification (imaging) =   1572)
Image Shift (X,Y):                                         .00,    .00

From Johnson-Matthey, Vacuum remelted, Batch BM1664
Optical emission: Al < 1ppm, Ca < 1 ppm,
Cr 2 ppm, Co 20 ppm, Cu 3 ppm, Ni 3 ppm
Si 60 ppm, Sn 10 ppm, Ag < 1 ppm
Oxygen 310 ppm, Nitrogen 10 ppm
Number of Data Lines:   3             Number of 'Good' Data Lines:   3
First/Last Date-Time: 08/13/2019 01:58:46 PM to 08/13/2019 02:07:32 PM
WARNING- Using Exponential Off-Peak correction for n ka
WARNING- Using Exponential Off-Peak correction for c ka
WARNING- Using Exponential Off-Peak correction for si ka
WARNING- Using Exponential Off-Peak correction for cu ka
WARNING- Using Exponential Off-Peak correction for o ka

Average Total Oxygen:         .000     Average Total Weight%:  113.447
Average Calculated Oxygen:    .000     Average Atomic Number:   24.060
Average Excess Oxygen:        .000     Average Atomic Weight:   41.316
Average ZAF Iteration:        5.00     Average Quant Iterate:     2.00

Using Conductive Coating Correction For Electron Absorption and X-Ray Transmission:
Sample Coating=C, Density=2.1 gm/cm3, Thickness=200 angstroms, Sin(Thickness)=311.145 angstroms

St  526 Set   3 Iron metal, Results in Elemental Weight Percents
 
ELEM:        N       C      Mo      Si      Ni       V      Cu       O      Fe      Cr      Mn
BGDS:      EXP     EXP     LIN     EXP     LIN     LIN     EXP     EXP     LIN     LIN     LIN
TIME:    40.00   40.00   40.00   40.00   20.00   20.00   20.00   40.00   20.00   20.00   20.00
BEAM:    49.56   49.56   49.56   49.56   49.56   49.56   49.56   49.56   49.56   49.56   49.56

ELEM:        N       C      Mo      Si      Ni       V      Cu       O      Fe      Cr      Mn   SUM 
   156   -.160  10.904   -.002    .011   -.006    .002   -.016    .357 102.982    .011    .008 114.091
   157   -.085  10.732   -.020    .007   -.019   -.013   -.003    .372 101.793   -.003   -.007 112.755
   158   -.285  10.792   -.025    .006    .011   -.002    .033    .432 102.560   -.019   -.009 113.495

AVER:    -.177  10.809   -.016    .008   -.004   -.004    .005    .387 102.445   -.003   -.003 113.447
SDEV:     .101    .087    .012    .002    .015    .007    .025    .040    .603    .015    .009    .669
SERR:     .058    .050    .007    .001    .008    .004    .015    .023    .348    .009    .005
%RSD:   -57.10     .81  -78.83   30.67 -328.34 -169.36  544.36   10.29     .59 -433.32 -346.16

PUBL:     n.a.    n.a.    n.a.    n.a.    n.a.    n.a.    n.a.    n.a. 100.000    n.a.    n.a. 100.000
%VAR:      ---     ---     ---     ---     ---     ---     ---     ---  (2.45)     ---     ---
DIFF:      ---     ---     ---     ---     ---     ---     ---     ---  (2.45)     ---     ---
STDS:      604     506     542     514     528     523     529     913     526     524     525

STKF:    .1637   .9635   .9910  1.0000  1.0000  1.0000   .9974   .2509  1.0000   .9988  1.0000
STCT:    28.99  853.49  258.45 1148.84  609.87  416.77  562.01  195.43  210.20  154.13  188.74

UNKF:   -.0006   .0285  -.0001   .0001   .0000  -.0001   .0000   .0021  1.0000   .0000   .0000
UNCT:     -.11   25.22    -.03     .06    -.03    -.02     .02    1.64  210.20    -.01     .00
UNBG:     1.31     .99     .39     .22    2.37     .64    2.90    2.40     .68     .28     .39

ZCOR:   2.8933  3.7968  1.1943  1.4322  1.0712   .8660  1.1135  1.8413  1.0245   .7501  1.0443
KRAW:   -.0037   .0295  -.0001   .0001   .0000  -.0001   .0000   .0084  1.0000   .0000   .0000
PKBG:      .92   26.40     .91    1.29     .99     .97    1.01    1.68  310.00     .98     .99
INT%:     ----    ----    ----    ----    ----    ----    ----    ----    ----    ----    -.04

Note that the total is quite high because the standard is carbon coated, and we are analyzing for carbon. And note also that the Fe concentration is wrong because we are adding 10 percent or so of carbon from the carbon coating.

But this extra carbon seen when a carbon coated standard is analyzed as an unknown, has zero effect on the standardization of unknowns because only the Fe intensities (in this standard) are being utilized for the unknown analysis. And of course we have specified in Probe for EPMA that the standards are carbon coated and the unknowns are not.

When we do this, we can analyze a uncoated sample using a standard that is carbon coated and the differences in the coating are compensated for in the matrix correction:

Un    6 H13 trav
TakeOff = 40.0  KiloVolt = 15.0  Beam Current = 50.0  Beam Size =    0
(Magnification (analytical) =  40000),        Beam Mode = Analog  Spot
(Magnification (default) =      400, Magnification (imaging) =   1572)
Image Shift (X,Y):                                         .00,    .00
Number of Data Lines:   5             Number of 'Good' Data Lines:   1
First/Last Date-Time: 08/13/2019 04:01:21 PM to 08/13/2019 04:27:59 PM
WARNING- Using Exponential Off-Peak correction for n ka
WARNING- Using Exponential Off-Peak correction for c ka
WARNING- Using Exponential Off-Peak correction for si ka
WARNING- Using Exponential Off-Peak correction for cu ka
WARNING- Using Exponential Off-Peak correction for o ka
WARNING- Using Time Dependent Intensity (TDI) Element Correction

Average Total Oxygen:         .000     Average Total Weight%:  100.766
Average Calculated Oxygen:    .000     Average Atomic Number:   24.805
Average Excess Oxygen:        .000     Average Atomic Weight:   46.435
Average ZAF Iteration:        4.00     Average Quant Iterate:     3.00

No Sample Coating and/or No Sample Coating Correction

Un    6 H13 trav, Results in Elemental Weight Percents
 
ELEM:        N       C      Mo      Si      Ni       V      Cu       O      Fe      Cr      Mn
BGDS:      EXP     EXP     LIN     EXP     LIN     LIN     EXP     EXP     LIN     LIN     LIN
TIME:    60.00   60.00   60.00   60.00   36.00   40.00   36.00   90.00   20.00   40.00   40.00
BEAM:    48.49   48.49   48.49   48.49   48.49   48.49   48.49   48.49   48.49   48.49   48.49

ELEM:        N       C      Mo      Si      Ni       V      Cu       O      Fe      Cr      Mn   SUM 
   165   3.406   2.406   1.224    .872    .194    .953    .092    .284  86.449   4.550    .334 100.766

AVER:    3.406   2.406   1.224    .872    .194    .953    .092    .284  86.449   4.550    .334 100.766
SDEV:     .000    .000    .000    .000    .000    .000    .000    .000    .000    .000    .000    .000
SERR:     .000    .000    .000    .000    .000    .000    .000    .000    .000    .000    .000
%RSD:      .00     .00     .00     .00     .00     .00     .00     .00     .00     .00     .00
STDS:      604     506     542     514     528     523     529     913     526     524     525

STKF:    .1637   .9635   .9910  1.0000  1.0000  1.0000   .9974   .2509  1.0000   .9988  1.0000
STCT:    29.07  853.49  258.45 1148.84  609.87  416.77  562.01  195.43  210.20  154.13  188.74

UNKF:    .0142   .0061   .0103   .0061   .0018   .0107   .0008   .0015   .8480   .0573   .0032
UNCT:     3.02    5.58    2.72    7.15    1.14    4.53     .48    1.28  181.90    9.01     .62
UNBG:     1.25     .74     .32     .28    2.39     .68    2.82    2.54     .64     .29     .39

ZCOR:   2.4014  3.9407  1.1922  1.4274  1.0606   .8939  1.1019  1.8968  1.0194   .7941  1.0359
KRAW:    .1040   .0065   .0105   .0062   .0019   .0109   .0009   .0066   .8654   .0585   .0033
PKBG:     3.41    8.50    9.52   26.38    1.48    7.63    1.17    1.51  283.94   32.20    2.60
INT%:     ----    ----    ----    ----    ----    ----    ----    ----     .00    ----   -1.01

TDI%:     .000   3.192   -.735    .000    .851    .000    .000  -4.725   2.630    .000    .000
DEV%:       .0     4.6     6.6      .0     4.9      .0      .0     3.8      .4      .0      .0
TDIF:     ---- LOG-LIN LOG-LIN    ---- LOG-LIN    ----    ---- LOG-LIN LOG-LIN    ----    ----
TDIT:      .00  111.00  112.00     .00  103.00     .00     .00  151.00   74.00     .00     .00
TDII:     ----    6.32    3.03    ----    3.55    ----    ----    3.69    183.    ----    ----
TDIL:     ----    1.84    1.11    ----    1.27    ----    ----    1.31    5.21    ----    ----