Interference corrections for combined WD-ED acquisitions

[David Steele]

Hi John,

I'm seeking advice on the best way to use PfE to perform an interference correction on a WD-acquired element by another element acquired by EDS!

The two scenarios of immediate relevance are: Fe L on F Ka and Si K on Ta Ma.

F and Ta (with other minors) are acquired by WD but I'd like to use the EDS to acquire Si and Fe (and a few other elements....)

Given that only the WDS-acquired (or specified/by stoichiometry) elements defined in the Elements/Cations window appear in the Standard Assignments window for use as interferences (EDS-assigned elements do not) I'm wondering how I can put the appropriate corrections for Si and Fe on Ta and F respectively in place.

The only way I can think of is to put Si and Fe in the Elements/Cations window and acquire WDS intensities as standards and unknowns as 'normal' but "DQ" (disable quantification) Si and Fe so that the ED intensities are used for the matrix corrections but the WD intensities of Si and Fe used for the interference corrections.  However this seems to defeat my intent which is to use EDS for the majors and leave the WDS for the minors/traces....

Perhaps I've hit on a couple of combinations that maybe the EDS should be left in the box and I should go with WDS-only??

Any thoughts?

(PS: I have tried the above and cutting the WD count times for Si and Fe to very minimal 1sec peaks and 0.5sec bgds but the counting stats/accuracy and precision of the interference corrections go "AWOL".....!)

Cheers,
David
QUT   

[John Donovan]

Hi John,

I'm seeking advice on the best way to use PfE to perform an interference correction on a WD-acquired element by another element acquired by EDS!

The two scenarios of immediate relevance are: Fe L on F Ka and Si K on Ta Ma.

F and Ta (with other minors) are acquired by WD but I'd like to use the EDS to acquire Si and Fe (and a few other elements....)

Given that only the WDS-acquired (or specified/by stoichiometry) elements defined in the Elements/Cations window appear in the Standard Assignments window for use as interferences (EDS-assigned elements do not) I'm wondering how I can put the appropriate corrections for Si and Fe on Ta and F respectively in place.

The only way I can think of is to put Si and Fe in the Elements/Cations window and acquire WDS intensities as standards and unknowns as 'normal' but "DQ" (disable quantification) Si and Fe so that the ED intensities are used for the matrix corrections but the WD intensities of Si and Fe used for the interference corrections.  However this seems to defeat my intent which is to use EDS for the majors and leave the WDS for the minors/traces....

Perhaps I've hit on a couple of combinations that maybe the EDS should be left in the box and I should go with WDS-only??

Any thoughts?

(PS: I have tried the above and cutting the WD count times for Si and Fe to very minimal 1sec peaks and 0.5sec bgds but the counting stats/accuracy and precision of the interference corrections go "AWOL".....!)

Cheers,
David
QUT

I've got a note from Julie C. on my stack that says: "Add EDS elements to interferences..."

A programmer's work is never done!

[John Donovan]

I think I've figured out a way to perform interference corrections on EDS *and* WDS data as requested by David Steele and Julie Chouinard... it won't be ready for another week or so, but here is how it will work:



Right now it's disabled of course, but eventually I will enable it and then the software will automatically convert your old EDS elements into the new method of integrated WDS-EDS.  This should provide the full array of data processing for both EDS and WDS elements- provided it applies of course!

The thing that I was stumped on, but I think I've figured out is that normally once the data is acquired you can't change (WDS) elements, but of course for EDS since the full EDS spectrum is stored you can.  So I had to figure out a way to allow that...

Stay tuned!

[John Donovan]

Thank-you David Steel and Julie Chouinard for this suggestion!

The latest version of Probe for EPMA now fully integrates WDS and EDS elements into a single array so that all analytical calculations can be performed on all elements regardless of their detector origin- with some exceptions (e.g., no TDI correction for EDS yet!).

However, the interference correction as requested by David and Julie can now be applied across all elements, whether EDS or WDS.  This is possible because the quantitative interference correction in PFE is based on concentration as opposed to intensity. So, because one can acquire their WDS elements along with a full EDS spectrum for every analysis, one can make fully quantitative interference corrections in PFE using elements by EDS added in off-line (or on-line) processing of data as seen here:



Note that the EDS elements can be changed or added post acquisition since we have a full EDS spectrum for every analysis!

Here is an example, Fe ka interfered by Mn. In this example I acquired Fe on the EDS spectrometer and Mn on the WDS spectrometer to make the interference a worst case situation as seen here:

St   25 Set   2 MnO synthetic
TakeOff = 40.0  KiloVolt = 15.0  Beam Current = 30.0  Beam Size =    0
(Magnification (analytical) =  40000),        Beam Mode = Analog  Spot
(Magnification (default) =      400, Magnification (imaging) =    800)
Image Shift (X,Y):                                          .00,   .00

Specimen from Michael Wittenauer (Purdue Univ.)
Starting mat'l 99.999%, SM # 317, 'skull melt' process
Mat. Res. Bull. 15, p 571, 1980
(possible intergrowths of Mn3O4 and small inclusions of Mn metal)
EPMA (UCB): SiO2=0.00, FeO=0.00, CaO=0.00, Al2O3=0
Number of Data Lines:   5             Number of 'Good' Data Lines:   5
First/Last Date-Time: 04/27/2015 12:17:14 PM to 04/27/2015 12:24:58 PM

Stage (or Beam Deflection) Coordinate Positions:
          X          Y          Z                 X          Y          Z
  131G   16350.00  -5011.000  -36.00000   132G   16354.00  -5011.000  -36.00000
  133G   16358.00  -5011.000  -36.00000   134G   16362.00  -5011.000  -36.00000
  135G   16366.00  -5011.000  -36.00000

Sample Coordinates Referenced to Fiducial Set  1 C:\UserData\StandardPOSData\alkali-glass_pos4.pos

On and Off Peak Positions:
ELEM:    si ka   mn ka   fe ka   sr la
ONPEAK 27889.0 23811.0 .000000 .000000
OFFSET -148.43 228.352 .000000 .000000
HIPEAK 28927.4 25932.2 .000000 .000000
LOPEAK 25547.2 23468.6 .000000 .000000
HI-OFF 1038.40 2121.17 .000000 .000000
LO-OFF -2341.8 -342.36 .000000 .000000

PHA Parameters:
ELEM:    si ka   mn ka   fe ka   sr la
DEAD:     2.90    3.31     .00     .00
BASE:      .56     .50     .00     .00
WINDOW    4.00    4.00     .00     .00
MODE:       -1      -1       0       0
GAIN:     782.    780.       .       .
BIAS:    1293.   1850.       .       .

Last (Current) On and Off Peak Count Times:
ELEM:    si ka   mn ka   fe ka   sr la
BGD:       OFF     OFF     EDS     EDS
BGDS:      EXP     EXP     EDS     EDS
SPEC:        4       3       0       0
CRYST:     TAP    LPET     EDS     EDS
ORDER:       1       1       0       0
ONTIM:   60.00   60.00   60.00   60.00
HITIM:   10.00   10.00    ----    ----
LOTIM:   10.00   10.00    ----    ----

Miscellaneous Sample Acquisition/Calculation Parameters:
KILO:    15.00   15.00   15.00   15.00
ENERGY   1.740   5.895   6.400   1.807
EDGE:    1.839   6.539   7.112   1.940
Eo/Ec:    8.16    2.29    2.11    7.73
STDS:       14      25     395     251

EDS Spectrum Parameters (for datarow=1):
Time Constant=     0, Preset Time=  60.0
Elapsed Time= 115.0, Dead Time(%)=  47.8
Live Time=  60.0,        Channels=  2048
Ev Per Chan=    10,  Max Counts= 1430951

On-Peak (off-peak corrected) or MAN On-Peak X-ray Counts (cps/1nA) (and Faraday/Absorbed Currents):
ELEM:    si ka   mn ka   fe ka   sr la   BEAM1   BEAM2
BGD:       OFF     OFF     EDS     EDS
SPEC:        4       3       0       0
CRYST:     TAP    LPET     EDS     EDS
ORDER:       1       1       0       0
  131G     .63 2063.85    4.43    1.19  29.800  29.800
  132G     .92 2062.26    4.82    1.53  29.798  29.797
  133G     .74 2063.35    5.11    1.10  29.797  29.798
  134G     .73 2062.81    5.39    1.53  29.798  29.797
  135G     .53 2064.64    5.38    1.49  29.798  29.797

AVER:      .71 2063.38    5.02    1.37  29.798  29.798
SDEV:      .15     .92     .41     .21    .001    .001
1SIG:      .04     .99     .05     .03
SIGR:     4.20     .94    7.67    7.50
SERR:      .07     .41     .18     .09
%RSD:    20.70     .04    8.09   15.18

Off-Peak (calculated) X-ray Counts (cps/1nA):
ELEM:    si ka   mn ka   fe ka   sr la
TYPE:  EXPONEN EXPONEN    ----    ----
COEF1:  6.0000  8.0000    ----    ----
COEF2:    ----    ----    ----    ----
COEF3:    ----    ----    ----    ----
  131G    1.57   34.05    ----    ----
  132G    1.45   34.32    ----    ----
  133G    1.52   33.97    ----    ----
  134G    1.45   33.63    ----    ----
  135G    1.48   33.94    ----    ----

Raw Hi-Peak X-ray Counts (cps/1nA):
ELEM:    si ka   mn ka   fe ka   sr la
  131G    1.48    8.34    ----    ----
  132G    1.37    8.22    ----    ----
  133G    1.44    8.38    ----    ----
  134G    1.33    8.77    ----    ----
  135G    1.39    8.38    ----    ----

Raw Lo-Peak X-ray Counts (cps/1nA):
ELEM:    si ka   mn ka   fe ka   sr la
  131G    1.85   42.97    ----    ----
  132G    1.70   43.47    ----    ----
  133G    1.77   42.83    ----    ----
  134G    1.83   42.01    ----    ----
  135G    1.77   42.78    ----    ----

We now specify the interference by simply clicking on Fe as seen here:



and selecting Mn as the interfering element, using MnO as the interference standard as seen here:



Now the interference correction is properly specified as seen here:



Naturally we'd probably want to measure trace Fe on the WDS spectrometer and Mn on the EDS spectrometer, but the correction works in either direction (and also between EDS elements!). Here are the results *without* the interference correction:

St   25 Set   2 MnO synthetic
TakeOff = 40.0  KiloVolt = 15.0  Beam Current = 30.0  Beam Size =    0
(Magnification (analytical) =  40000),        Beam Mode = Analog  Spot
(Magnification (default) =      400, Magnification (imaging) =    800)
Image Shift (X,Y):                                          .00,   .00

Specimen from Michael Wittenauer (Purdue Univ.)
Starting mat'l 99.999%, SM # 317, 'skull melt' process
Mat. Res. Bull. 15, p 571, 1980
(possible intergrowths of Mn3O4 and small inclusions of Mn metal)
EPMA (UCB): SiO2=0.00, FeO=0.00, CaO=0.00, Al2O3=0
Number of Data Lines:   5             Number of 'Good' Data Lines:   5
First/Last Date-Time: 04/27/2015 12:17:14 PM to 04/27/2015 12:24:58 PM
WARNING- Using Exponential Off-Peak correction for si ka
WARNING- Using Exponential Off-Peak correction for mn ka

Average Total Oxygen:         .000     Average Total Weight%:  100.966
Average Calculated Oxygen:    .000     Average Atomic Number:   21.217
Average Excess Oxygen:        .000     Average Atomic Weight:   35.590
Average ZAF Iteration:        3.00     Average Quant Iterate:     2.00

St   25 Set   2 MnO synthetic, Results in Elemental Weight Percents
 
ELEM:       Si      Mn      Fe      Sr       O
TYPE:     ANAL    ANAL    ANAL    ANAL    SPEC
BGDS:      EXP     EXP     EDS     EDS
TIME:    60.00   60.00   60.00   60.00     ---
BEAM:    29.80   29.80   29.80   29.80     ---

ELEM:       Si      Mn      Fe      Sr       O   SUM 
   131    .045  77.378    .775    .079  22.554 100.830
   132    .067  77.343    .844    .101  22.554 100.908
   133    .054  77.406    .893    .073  22.554 100.979
   134    .053  77.383    .942    .101  22.554 101.033
   135    .038  77.448    .940    .099  22.554 101.079

AVER:     .051  77.392    .879    .090  22.554 100.966
SDEV:     .011    .039    .071    .014    .000    .099
SERR:     .005    .017    .032    .006    .000
%RSD:    20.69     .05    8.07   15.14     .00

PUBL:     n.a.  77.446    n.a.    n.a.  22.554 100.000
%VAR:      ---  (-.07)     ---     ---     .00
DIFF:      ---  (-.05)     ---     ---    .000
STDS:       14      25     395     251     ---

STKF:    .4101   .7341   .6779   .4268     ---
STCT:   765.85 2063.82  401.70  866.41     ---

UNKF:    .0004   .7340   .0085   .0007     ---
UNCT:      .71 2063.38    5.02    1.37     ---
UNBG:     1.49   33.98     .00     .00     ---

ZCOR:   1.3444  1.0544  1.0363  1.3427     ---
KRAW:    .0009   .9998   .0125   .0016     ---
PKBG:     1.48   61.72     .00     .00     ---
INT%:     ----    ----    ----    ----     ---

Note the roughly 0.9 wt% apparent Fe in MnO.  Now *with* the interference correction:

St   25 Set   2 MnO synthetic
TakeOff = 40.0  KiloVolt = 15.0  Beam Current = 30.0  Beam Size =    0
(Magnification (analytical) =  40000),        Beam Mode = Analog  Spot
(Magnification (default) =      400, Magnification (imaging) =    800)
Image Shift (X,Y):                                          .00,   .00

Specimen from Michael Wittenauer (Purdue Univ.)
Starting mat'l 99.999%, SM # 317, 'skull melt' process
Mat. Res. Bull. 15, p 571, 1980
(possible intergrowths of Mn3O4 and small inclusions of Mn metal)
EPMA (UCB): SiO2=0.00, FeO=0.00, CaO=0.00, Al2O3=0
Number of Data Lines:   5             Number of 'Good' Data Lines:   5
First/Last Date-Time: 04/27/2015 12:17:14 PM to 04/27/2015 12:24:58 PM
WARNING- Using Exponential Off-Peak correction for si ka
WARNING- Using Exponential Off-Peak correction for mn ka

Average Total Oxygen:         .000     Average Total Weight%:  100.124
Average Calculated Oxygen:    .000     Average Atomic Number:   21.177
Average Excess Oxygen:        .000     Average Atomic Weight:   35.482
Average ZAF Iteration:        3.00     Average Quant Iterate:     3.00

St   25 Set   2 MnO synthetic, Results in Elemental Weight Percents
 
ELEM:       Si      Mn      Fe      Sr       O
TYPE:     ANAL    ANAL    ANAL    ANAL    SPEC
BGDS:      EXP     EXP     EDS     EDS
TIME:    60.00   60.00   60.00   60.00     ---
BEAM:    29.80   29.80   29.80   29.80     ---

ELEM:       Si      Mn      Fe      Sr       O   SUM 
   131    .045  77.415   -.106    .079  22.554  99.987
   132    .067  77.380   -.035    .101  22.554 100.066
   133    .054  77.443    .014    .072  22.554 100.137
   134    .053  77.421    .064    .101  22.554 100.192
   135    .038  77.485    .061    .099  22.554 100.237

AVER:     .051  77.429    .000    .090  22.554 100.124
SDEV:     .011    .039    .071    .014    .000    .100
SERR:     .005    .017    .032    .006    .000
%RSD:    20.69     .05    ----   15.14     .00

PUBL:     n.a.  77.446    n.a.    n.a.  22.554 100.000
%VAR:      ---  (-.02)     ---     ---     .00
DIFF:      ---  (-.02)     ---     ---    .000
STDS:       14      25     395     251     ---

STKF:    .4101   .7341   .6779   .4268     ---
STCT:   765.85 2063.82  401.69  866.41     ---

UNKF:    .0004   .7340   .0000   .0007     ---
UNCT:      .71 2063.38     .00    1.37     ---
UNBG:     1.49   33.98     .00     .00     ---

ZCOR:   1.3434  1.0549  1.0366  1.3418     ---
KRAW:    .0009   .9998   .0000   .0016     ---
PKBG:     1.48   61.72     .00     .00     ---
INT%:     ----    ---- -100.58    ----     ---

Holy cow, that is pretty cool!

I should also mention that this new capability is going to open up all sorts of flexibility for the analyst and I'm still sorting out this in my head, for example the Evaluate application which allows one to check standards for self-consistency, will now work with elements regardless whether they were acquired using WDS or EDS!

[Probeman]

Some additional thoughts: both the blank correction and halogen correction can be applied to EDS elements with this new feature...

Finally here is the original EDS-WDS integration topic for further discussion:

https://probesoftware.com/smf/index.php?topic=79.0

[sckuehn]

John,

I really like this new approach with WDS and EDS now more unified rather than being dealt with in two different areas of the application. Too bad I have an older instrument not able to run the newer software.

However, it sounds like I could use the new version for offline processing even if I am still using the old version for acquisition. I am guessing that I would have to keep separate copies of the mdb, though, because the conversion may make the files incompatible with older software.

Any chance that the changes might make it easier to implement EDS TDI too?

- Steve

[John Donovan]

I really like this new approach with WDS and EDS now more unified rather than being dealt with in two different areas of the application. Too bad I have an older instrument not able to run the newer software.

Hi Steve,
I would look for a lab replacing a JEOL 8900 or Cameca SX100.  That latter is what Univ. of Oklahoma did.  You might ask George Morgan what he paid for his instrument from Sandia.

However, it sounds like I could use the new version for offline processing even if I am still using the old version for acquisition. I am guessing that I would have to keep separate copies of the mdb, though, because the conversion may make the files incompatible with older software.

I can't remember if you have all tunable spectrometers or some fixed spectrometers also on your SEMQ instrument.  If you have only tunable spectrometers, I think it is very likely you could reprocess data in the latest version of Probe for EPMA, but you should test it and see.  It might even work if you have some fixed spectrometers but there might be some complaints about zero spectrometer numbers for those channels...

Any chance that the changes might make it easier to implement EDS TDI too?

Not quite sure what you are asking here.  Do you mean sub interval counting for EDS spectrum acquisition?  If so, then that is an idea that hasn't been implemented yet.  I would probably just say: acquire the beam sensitive elements with WDS and the other elements with EDS.

My next idea for EDS WDS integration is here (see the 4th item):

https://probesoftware.com/smf/index.php?topic=400.msg3702#msg3702

It will make existing WDS EDS ROI mapping methods obsolete.
john

[sckuehn]

I'd be happy to give post-processing on the newer software a try. I have an application that could use the EDS Si for an interference correction.

Concord's ARL instrument has 4 tunable plus a Bruker SDD EDS, although we have an opportunity to obtain a bunch of additional spectrometers in the near future. This could be configured as 2 additional tunable units or 2 fixed (for 4 elements - 2 each, MAN required), or some combination, plus extras for teaching demonstrations or possible application-specific re-configuration. Eight elements simultaneously with MAN backgrounds would do great things for precision and acquisition time.

I've been entertaining the idea of a newer instrument on and off. Unlocking the newer PFE software features would be great! On the other hand, the ARL is robust, stable, and relatively cheap to keep running. We also made a major investment in a new EDS only a few years ago, and that may not be so easily transferred to a different instrument.

For EDS TDI, sub interval counting just like with WDS is exactly what I'm thinking. You may recall me asking about this 2-4 years ago.  I use the WDS TDI all the time.

[John Donovan]

Concord's ARL instrument has 4 tunable plus a Bruker SDD EDS, although we have an opportunity to obtain a bunch of additional spectrometers in the near future. This could be configured as 2 additional tunable units or 2 fixed (for 4 elements - 2 each, MAN required), or some combination, plus extras for teaching demonstrations or possible application-specific re-configuration. Eight elements simultaneously with MAN backgrounds would do great things for precision and acquisition time.

Hi Steve,
My SEMQ at UC Berkeley (25 years ago) had 4 tunable and 4 fixed spectrometers and could acquire 8 elements in 10 seconds using MAN backgrounds!

Reviewers could never seem to understand how we were able to do this... but it made acquisition on beam sensitive glasses very nice! 

I dream of a new six (or eight!) tunable spectrometer instrument from Cameca or JEOL all the time...
john