Author Topic: Time Dependent Intensity (TDI) Corrections (Read 29541 times)

John Donovan · « **on:** July 05, 2013, 09:34:51 AM »

A recent modification in v. 10.0.4 of PFE allows the user to review the TDI intensity data for *all* samples, which have TDI data acquired, from the Run | Display Time Dependent (TDI) and Alternating (on/off) Intensities menu.

The change now allows the user to see intensity data for TDI channels even if those data points have *not* been assigned the TDI correction, and in addition, the program now adds the string "(TDI off) to those data points to indicate that they are not using the TDI correction, though TDI data was acquired and is available for assignment as seen here:

John Donovan · « **Reply #1 on:** October 01, 2013, 04:16:59 PM »

The Time Dependent Intensity (TDI) correction for changes in x-ray intensity as a function of acquisition time is well known though not completely understood. The physics of ion migration/volatilization as a function of thermal conductivity, ion mobility and electron dose is complex and very likely convolves together several different physical processes with different time scales.

But for the purposes of correcting for this artifact since the sample is what it is, we are left with reducing beam current, increasing beam diameter, decreasing beam exposure (acquisition) time, cooling the sample cryogenically or some combination of these procedures along with a software correction for these effects which, depending on the element, can result in decreases or increases in x-ray intensity over time.

A typical hydrous glass with a normal (log-linear) TDI correction for Na, Si, Al (and Ca and Ti) is shown here:

ELEM: Na Si K Al Mg Ca Ti Mn Fe P Cr O H TYPE: ANAL ANAL ANAL ANAL ANAL ANAL ANAL ANAL ANAL ANAL ANAL CALC SPEC BGDS: MAN MAN LIN MAN MAN MAN LIN LIN MAN EXP LIN TIME: 80.00 30.00 40.00 40.00 60.00 80.00 20.00 20.00 60.00 60.00 30.00 BEAM: 20.08 20.08 20.08 20.08 20.08 20.08 20.08 20.08 20.08 20.08 20.08 ELEM: Na Si K Al Mg Ca Ti Mn Fe P Cr O H SUM 235 .315 26.882 .449 4.251 5.416 4.877 .925 .343 11.099 .199 .055 44.303 .000 99.113 236 .487 26.014 .526 4.256 1.367 5.613 .932 .344 16.725 .194 .027 42.621 .000 99.107 237 .396 26.547 .455 4.244 4.251 5.551 .907 .344 12.211 .192 .037 43.735 .000 98.870 238 .359 26.017 .452 4.177 2.586 5.918 .892 .359 14.556 .195 .033 42.777 .000 98.320 AVER: .389 26.365 .471 4.232 3.405 5.490 .914 .348 13.648 .195 .038 43.359 .000 98.852 SDEV: .073 .426 .037 .037 1.787 .439 .018 .008 2.507 .003 .012 .799 .000 .373 SERR: .037 .213 .018 .018 .894 .220 .009 .004 1.254 .001 .006 .400 .000 %RSD: 18.77 1.62 7.83 .87 52.49 8.00 2.01 2.18 18.37 1.49 31.81 1.84 .00 TDI%: 27.702 -.037 ---- -.154 ---- -1.026 .563 ---- ---- ---- ---- ---- ---- DEV%: 10.3 .6 ---- 1.1 ---- 2.9 8.2 ---- ---- ---- ---- ---- ---- TDIF: LINEAR LINEAR ---- LINEAR ---- LINEAR LINEAR ---- ---- ---- ---- ---- ---- TDIT: 97.00 48.75 ---- 58.00 ---- 98.75 37.00 ---- ---- ---- ---- ---- ---- TDII: 72.0 10400. ---- 1897. ---- 169. 98.3 ---- ---- ---- ---- ---- ----

As we can see, the TDI % log-linear correction above highlighted in red, is approximately 28% which seems reasonable until one examines the actual extrapolation to zero time, and then a problem is obvious:

Clearly the intensity change over time is more than a simple exponential.

However, by fitting the log intensity to a quadratic and extrapolating again, we obtain the following results as shown below:

ELEM: Na Si K Al Mg Ca Ti Mn Fe P Cr O H TYPE: ANAL ANAL ANAL ANAL ANAL ANAL ANAL ANAL ANAL ANAL ANAL CALC SPEC BGDS: MAN MAN LIN MAN MAN MAN LIN LIN MAN EXP LIN TIME: 80.00 30.00 40.00 40.00 60.00 80.00 20.00 20.00 60.00 60.00 30.00 BEAM: 20.08 20.08 20.08 20.08 20.08 20.08 20.08 20.08 20.08 20.08 20.08 ELEM: Na Si K Al Mg Ca Ti Mn Fe P Cr O H TYPE: ANAL ANAL ANAL ANAL ANAL ANAL ANAL ANAL ANAL ANAL ANAL CALC SPEC BGDS: MAN MAN LIN MAN MAN MAN LIN LIN MAN EXP LIN TIME: 80.00 30.00 40.00 40.00 60.00 80.00 20.00 20.00 60.00 60.00 30.00 BEAM: 20.08 20.08 20.08 20.08 20.08 20.08 20.08 20.08 20.08 20.08 20.08 ELEM: Na Si K Al Mg Ca Ti Mn Fe P Cr O H SUM 235 .436 26.890 .449 4.253 5.421 4.876 .925 .343 11.099 .199 .055 44.359 .000 99.306 236 .584 26.021 .526 4.258 1.368 5.613 .932 .344 16.725 .194 .027 42.665 .000 99.257 237 .461 26.551 .455 4.245 4.253 5.551 .907 .344 12.211 .192 .037 43.766 .000 98.973 238 .450 26.023 .452 4.179 2.588 5.918 .892 .359 14.556 .195 .033 42.818 .000 98.462 AVER: .483 26.371 .471 4.234 3.407 5.490 .914 .348 13.648 .195 .038 43.402 .000 99.000 SDEV: .068 .427 .037 .037 1.789 .439 .018 .008 2.507 .003 .012 .803 .000 .387 SERR: .034 .213 .018 .018 .894 .220 .009 .004 1.254 .001 .006 .401 .000 %RSD: 14.19 1.62 7.83 .87 52.50 8.00 2.01 2.18 18.37 1.49 31.81 1.85 .00 TDI%: 54.831 -.037 ---- -.154 ---- -1.026 .563 ---- ---- ---- ---- ---- ---- DEV%: 5.9 .6 ---- 1.1 ---- 2.9 8.2 ---- ---- ---- ---- ---- ---- TDIF: QUADRA LINEAR ---- LINEAR ---- LINEAR LINEAR ---- ---- ---- ---- ---- ---- TDIT: 97.00 48.75 ---- 58.00 ---- 98.75 37.00 ---- ---- ---- ---- ---- ---- TDII: 87.2 10400. ---- 1897. ---- 169. 98.3 ---- ---- ---- ---- ---- ----

This was accomplished by merely regressing with the "hyper-exponential" fit shown here in the Standard Assignments dialog for Na:

Therefore, one may improve the accuracy of the TDI correction considerably without having to reduce beam current or increase beam diameter, as shown in the "hyper-exponential" fit below:

John Donovan · « **Reply #2 on:** October 01, 2013, 04:55:54 PM »

The importance of the matrix correction in these types of beam sensitive samples (e.g., hydrated) may be illustrated with the following data:

Results in Oxide Weight Percents ELEM: CaO K2O FeO SiO2 MgO Na2O Al2O3 TiO2 P2O5 O H2O SUM 58 .411 4.900 .537 74.603 .036 2.644 11.975 .073 .000 .000 .000 95.180 59 .437 4.624 .612 75.386 .033 1.841 12.024 .113 .000 .000 .000 95.070 60 .423 4.998 .526 74.715 .023 2.765 12.113 .065 .000 .000 .000 95.628 61 .423 4.853 .637 75.000 .047 2.137 12.089 .055 .000 .000 .000 95.240 62 .435 4.847 .677 74.613 .048 2.561 12.089 .036 .011 .000 .000 95.317 AVER: .426 4.844 .598 74.863 .037 2.390 12.058 .068 .002 .000 .000 95.287 SDEV: .011 .137 .065 .333 .010 .387 .057 .028 .005 .000 .000 SERR: .005 .061 .029 .149 .005 .173 .025 .013 .002 .000 .000 %RSD: 2.5 2.8 10.8 .4 27.9 16.2 .5 41.5 223.6 418.3 .0 ZCOR: 1.1202 1.1536 1.1987 1.2061 1.4303 1.8227 1.2568 1.1983 1.4641 .0000 .0000 KRAW: .0161 .3080 .0057 1.0344 .0003 .1323 .7242 .0006 -.0004 .0000 .0000 PKBG: 4.96 59.36 4.84 370.59 1.42 30.93 77.40 1.46 .87 .00 .00
The above analysis was performed without a TDI correction and also without specifying water in the matrix correction. The water by difference calculation yields approximately 4.7 wt% H2O. Note the average SiO2 and Na2O concentrations highlighted in red for comparison with the results below.

Next we specify the TDI correction as described above and obtain the results shown here:

Results in Oxide Weight Percents ELEM: CaO K2O FeO SiO2 MgO Na2O Al2O3 TiO2 P2O5 O H2O SUM 58 .411 4.867 .537 74.378 .037 3.626 12.035 .030 .000 .000 .000 95.921 59 .437 4.740 .612 75.016 .034 2.426 12.063 .194 .000 .000 .000 95.521 60 .423 4.933 .526 74.057 .024 3.805 12.178 .076 .000 .000 .000 96.022 61 .423 5.220 .636 74.552 .048 3.025 12.147 .077 .000 .000 .000 96.128 62 .435 4.715 .677 74.794 .049 3.503 12.146 .048 .011 .000 .000 96.378 AVER: .426 4.895 .598 74.559 .038 3.277 12.114 .085 .002 .000 .000 95.994 SDEV: .011 .203 .065 .370 .010 .557 .062 .064 .005 .000 .000 SERR: .005 .091 .029 .166 .005 .249 .028 .029 .002 .000 .000 %RSD: 2.5 4.1 10.8 .5 27.3 17.0 .5 75.6 223.6 .0 .0 ZCOR: 1.1198 1.1530 1.1983 1.2092 1.4391 1.8180 1.2625 1.1979 1.4627 .0000 .0000 KRAW: .0161 .3114 .0057 1.0276 .0003 .1819 .7242 .0008 -.0004 .0000 .0000 PKBG: 4.97 60.00 4.84 369.28 1.43 42.09 77.77 1.58 .87 .00 .00 INT%: .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 VOL%: .000 1.135 .000 -.658 .000 36.111 .000 20.283 .000 .000 .000 DEV%: .0 .2 .0 .0 .0 .4 .0 20.3 .0 .0 .0
The above analysis was performed *with* the TDI correction, but still without specifying water in the matrix correction. The water by difference calculation now yields approximately 4.0 wt% H2O. Note also that due to the TDI correction, the SiO2 concentration has decreased slightly and the Na2O concentration has increased significantly as highlighted in red.

Finally, we specify water in the matrix correction and obtain the results shown here:

Results in Oxide Weight Percents ELEM: CaO K2O FeO SiO2 MgO Na2O Al2O3 TiO2 P2O5 O H2O SUM 58 .414 4.885 .542 74.776 .039 3.668 12.139 .030 .000 .000 3.506 100.000 59 .441 4.759 .618 75.464 .036 2.457 12.181 .195 .000 .000 3.849 100.000 60 .426 4.951 .531 74.441 .026 3.848 12.281 .076 .000 .000 3.420 100.000 61 .426 5.239 .641 74.934 .050 3.058 12.249 .077 .000 .000 3.327 100.000 62 .438 4.730 .682 75.149 .051 3.539 12.240 .048 .011 .000 3.112 100.000 AVER: .429 4.913 .603 74.953 .041 3.314 12.218 .085 .002 .000 3.443 100.000 SDEV: .011 .203 .065 .385 .010 .562 .057 .065 .005 .000 .270 SERR: .005 .091 .029 .172 .005 .251 .025 .029 .002 .000 .121 %RSD: 2.5 4.1 10.8 .5 25.7 16.9 .5 75.6 223.6 -104.6 7.9 ZCOR: 1.1241 1.1571 1.2053 1.2156 1.4515 1.8372 1.2731 1.2037 1.4591 .0000 .0000 KRAW: .0161 .3114 .0057 1.0276 .0004 .1821 .7244 .0008 -.0004 .0000 .0000 PKBG: 5.03 60.00 4.89 372.20 1.46 43.35 78.99 1.58 .87 .00 .00 INT%: .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 VOL%: .000 1.135 .000 -.658 .000 36.111 .000 20.283 .000 .000 .000 DEV%: .0 .2 .0 .0 .0 .4 .0 20.3 .0 .0 .0
It is quite interesting to note that by simply adding water to the matrix correction, the concentration of the Si is significantly increased resulting in a water by difference calculation that now yields approximately 3.4 wt% H2O. Note that both the average SiO2 and Na2O concentrations have increased as seen highlighted in red.

This is due to the fact that water (mostly oxygen) absorbs Si Ka (and most of the other emission lines here) more strongly than Si, thus resulting in larger Si (and other elements) and smaller H2O values when the matrix effects of H2O are included in the matrix correction.

John Donovan · « **Reply #3 on:** October 04, 2013, 09:47:37 AM »

Don't forget your beam sensitive standards!

By checking the little box here in the Special Options dialog (from Acquire!) (thanks to Paul Carpenter for the idea):

you can also perform TDI corrections on your beam sensitive standards. E.g., anhydrite, apatite, etc:

However, note that Ca here could be perhaps better fit using the quadratic (hyper-exponential) fit as seen here:

Also, if you are measuring oxygen, this element intensity is often affected by beam damage:

Happy probing!

Richard Walshaw · « **Reply #4 on:** February 19, 2014, 09:04:40 AM »

Hi All

I'm looking through the literature to get a feel for who's done what using the TDI correction in PfEPMA.
Can you please point me toward any recent (not old) publications you know about or have written.

Thanks

Richard

Probeman · « **Reply #5 on:** February 19, 2014, 12:32:35 PM »

Quote from: Richard Walshaw on February 19, 2014, 09:04:40 AM

I'm looking through the literature to get a feel for who's done what using the TDI correction in PfEPMA.
Can you please point me toward any recent (not old) publications you know about or have written.

I have to issue a big apology here.

I have been swamped these last few years with other work and have not taken the time to properly document the new TDI methods I've developed recently and which are still on going...

Normally, I try to publish everything that is worthwhile, primarily so it is scientifically documented, but also so every feature in the software gets its proper "bona fides" so to speak!

I have asked a couple of younger colleagues to help me write up these TDI methods into a full paper but that process has only just got going last year, though we hope to have something finished this year.

In the meantime I have some links to abstracts here:

Glass abstract with Michael Rowe:
http://epmalab.uoregon.edu/pdfs/gold2005_Donovan.pdf

The presentation for the above abstract:
http://epmalab.uoregon.edu/reports/Improving%20Glass%20Analyses.pdf

A white paper on TDI effects when measuring Na, Si and O in hydrous glasses (method of B. Nash, et al.):
http://epmalab.uoregon.edu/reports/Withers%20hydrous%20glass.pdf

Probeman · « **Reply #6 on:** March 10, 2014, 04:21:01 PM »

Here is some additional explanation of the log window output for TDI sample quantitative correction results:

Un 27 MAM IW2 C4-ext-1, Results in Elemental Weight Percents ELEM: Na Si K Al Mg Ca Ti Mn Fe P Cr O H TYPE: ANAL ANAL ANAL ANAL ANAL ANAL ANAL ANAL ANAL ANAL ANAL CALC SPEC BGDS: MAN MAN LIN MAN MAN MAN LIN LIN MAN EXP LIN TIME: 79.98 29.98 40.00 40.00 60.00 80.00 20.00 20.00 60.00 60.00 30.00 BEAM: 20.09 20.09 20.09 20.09 20.09 20.09 20.09 20.09 20.09 20.09 20.09 ELEM: Na Si K Al Mg Ca Ti Mn Fe P Cr O H SUM 223 .507 25.693 .391 4.354 4.390 5.922 1.025 .385 12.555 .207 .024 43.328 .000 98.780 224 .256 25.891 .329 4.272 4.917 5.856 .863 .372 10.511 .181 .032 42.976 .000 96.457 225 .333 26.002 .390 4.423 3.930 6.318 1.008 .340 12.968 .196 .032 43.620 .000 99.559 226 .351 25.804 .379 4.197 4.793 5.837 .954 .338 12.482 .191 .055 43.403 .000 98.784 227 .289 26.034 .374 4.350 5.419 5.865 .955 .338 11.326 .193 .046 43.869 .000 99.057 228 .308 25.807 .365 4.239 4.010 6.126 .903 .333 13.373 .186 .037 43.230 .000 98.918 AVER: .341 25.872 .371 4.306 4.577 5.987 .951 .351 12.203 .193 .037 43.404 .000 98.593 SDEV: .088 .130 .023 .084 .574 .194 .061 .022 1.076 .009 .011 .310 .000 1.085 SERR: .036 .053 .009 .034 .234 .079 .025 .009 .439 .004 .005 .127 .000 %RSD: 25.76 .50 6.16 1.95 12.54 3.24 6.43 6.15 8.82 4.60 29.73 .72 .00 STDS: 336 162 374 336 162 162 22 25 162 285 396 0 0 STKF: .0735 .2018 .1132 .1331 .0568 .1027 .5547 .7341 .0950 .1599 .3050 .0000 .0000 STCT: 2515.7 10026.0 6009.1 8230.4 2834.3 337.0 6393.2 14939.3 599.4 9652.2 5184.8 .0 .0 UNKF: .0016 .2045 .0034 .0306 .0293 .0557 .0082 .0029 .1039 .0014 .0003 .0000 .0000 UNCT: 55.4 10158.6 178.3 1893.2 1460.9 182.6 94.5 59.8 655.3 82.5 5.6 .0 .0 UNBG: 10.4 9.8 30.1 28.5 19.0 1.5 7.6 19.2 8.0 39.1 14.8 .0 .0 ZCOR: 2.1042 1.2653 1.1057 1.4062 1.5655 1.0759 1.1608 1.1946 1.1745 1.4095 1.1305 .0000 .0000 KRAW: .0220 1.0132 .0297 .2300 .5154 .5420 .0148 .0040 1.0933 .0085 .0011 .0000 .0000 PKBG: 6.34 1039.69 6.95 67.51 77.70 126.12 13.71 4.12 82.50 3.12 1.39 .00 .00 INT%: ---- ---- ---- ---- -.11 ---- ---- ---- .00 ---- ---- ---- ---- TDI%: 39.559 -.098 ---- -.185 ---- -1.692 -.502 ---- ---- ---- ---- ---- ---- DEV%: 4.8 .6 ---- 1.1 ---- 2.6 7.5 ---- ---- ---- ---- ---- ---- TDIF: QUADRA LINEAR ---- LINEAR ---- LINEAR LINEAR ---- ---- ---- ---- ---- ---- TDIT: 98.33 49.83 ---- 58.33 ---- 98.33 37.83 ---- ---- ---- ---- ---- ---- TDII: 64.7 10169. ---- 1921. ---- 184. 102. ---- ---- ---- ---- ---- ----

Here is a "key":

John Donovan · « **Reply #7 on:** June 11, 2014, 12:20:03 PM »

Here's a TDI plot for some points in a high crystallized basalt where the student was trying to find the residual glass chemistry, but often hit some crystals.

Even though the chemistry of the two phases are very similar, the glass compositions are easily distinguished from the crystal compositions:

Gareth D Hatton · « **Reply #8 on:** June 12, 2014, 06:56:48 AM »

That is very nice

Probeman · « **Reply #9 on:** December 18, 2014, 12:38:36 PM »

This is an "extreme" situation, but if you ever have to characterize a Na rich glass such as NIST K-375 (which would never occur naturally I think!), the TDI correction in PFE will come to your rescue. In this case even the linear and hyper exponential TDI regressions will under fit this type of Na loss over time.

But the double exponential TDI correction seen here:

can offer some help, though this is a scary extrapolation for sure as you can see:

Basically in this case, one is losing over 50% of the Na intensity in less than 8 seconds after the faraday cup is removed!

How accurate is the extrapolation this particular case? I hesitate to reveal, but.... all in the name of science, here it is:

St 173 Set 24 K-0375 NBS glass TakeOff = 40.0 KiloVolt = 15.0 Beam Current = 100. Beam Size = 10 (Magnification (analytical) = 40000), Beam Mode = Analog Spot (Magnification (default) = 400, Magnification (imaging) = 800) Image Shift (X,Y): .00, .00 Number of Data Lines: 5 Number of 'Good' Data Lines: 5 First/Last Date-Time: 05/10/2012 02:30:56 AM to 05/10/2012 02:38:17 AM WARNING- Using Exponential Off-Peak correction for si ka WARNING- Using Exponential Off-Peak correction for p ka WARNING- Using Time Dependent Intensity (TDI) Element Correction WARNING- Using Time Dependent Intensity (TDI) Weighting Factor of 10 Average Total Oxygen: .000 Average Total Weight%: 98.935 Average Calculated Oxygen: .000 Average Atomic Number: 16.454 Average Excess Oxygen: .000 Average Atomic Weight: 22.833 Average ZAF Iteration: 4.00 Average Quant Iterate: 2.00 St 173 Set 24 K-0375 NBS glass, Results in Elemental Weight Percents ELEM: Na Si Ca Fe P Zn Ba U O TYPE: ANAL ANAL ANAL ANAL ANAL SPEC SPEC SPEC SPEC BGDS: LIN EXP LIN LIN EXP TIME: 20.00 20.00 20.00 20.00 20.00 BEAM: 100.72 100.72 100.72 100.72 100.72 ELEM: Na Si Ca Fe P Zn Ba U O SUM 371 9.539 32.765 .001 .007 .004 4.940 10.370 .110 42.320 100.056 372 8.727 32.573 .005 -.008 .012 4.940 10.370 .110 42.320 99.049 373 9.551 32.345 .008 -.014 .014 4.940 10.370 .110 42.320 99.644 374 8.440 32.289 .001 -.029 .008 4.940 10.370 .110 42.320 98.450 375 7.829 31.913 .001 .013 -.021 4.940 10.370 .110 42.320 97.476 AVER: 8.817 32.377 .003 -.006 .003 4.940 10.370 .110 42.320 98.935 SDEV: .739 .321 .003 .017 .014 .000 .000 .000 .000 1.017 SERR: .331 .144 .001 .008 .006 .000 .000 .000 .000 %RSD: 8.38 .99 89.41 -273.00 412.33 .00 .00 .00 .00 PUBL: 10.420 31.830 n.a. n.a. n.a. 4.940 10.370 .110 42.320 99.990 %VAR: -15.38 1.72 --- --- --- .00 .00 .00 .00 DIFF: -1.603 .547 --- --- --- .000 .000 .000 .000 STDS: 336 14 285 162 285 0 0 0 0 STKF: .0735 .4101 .3596 .0950 .1599 .0000 .0000 .0000 .0000 STCT: 78.78 80.11 605.49 66.49 40.15 .00 .00 .00 .00 UNKF: .0427 .2552 .0000 -.0001 .0000 .0000 .0000 .0000 .0000 UNCT: 45.72 49.86 .05 -.04 .01 .00 .00 .00 .00 UNBG: 3.34 .16 1.23 .99 .05 .00 .00 .00 .00 ZCOR: 2.0679 1.2685 1.0646 1.1515 1.4651 .0000 .0000 .0000 .0000 KRAW: .5803 .6224 .0001 -.0006 .0001 .0000 .0000 .0000 .0000 PKBG: 14.69 308.09 1.04 .97 1.26 .00 .00 .00 .00 TDI%: 4918.968 -8.230 1.354 ---- ---- ---- ---- ---- ---- DEV%: 9.8 .4 32.9 ---- ---- ---- ---- ---- ---- TDIF: LOG-LOG LOG-LOG LOG-LIN ---- ---- ---- ---- ---- ---- TDIT: 74.20 74.40 71.80 ---- ---- ---- ---- ---- ---- TDII: 45.6 49.9 1.29 ---- ---- ---- ---- ---- ----

So, ok we are off by some 15% relative error for Na, *but* the TDI% correction is almost 5000%. That is not a typo!

And... we are *still* under fitting the data! How about a triple exponential, anyone?

That is what happens when you use a 100 nA beam on such a beam sensitive material...

JohnF · « **Reply #10 on:** January 06, 2015, 09:58:01 AM »

I am trying to toggle the TDI on and off to see "what if". I am using the 9.6.3 PfE flavor on our SX51 which will continue operation for several more years. Regardless of _3_ ways of trying to turn off the TDI to see what the integrated count would be, nothing seems to work. The 3 ways are (1) in the Analytical/Analysis Options, uncheck the Use assigned or Self TDI corrections on Unknown box, (2) Assignment properties, click the No TDI Califbration Correction radio button, (3) under Standard and Interference Assignments, click the Remove TDI Correction. Have other people successfully been able to 'turn off' and see the difference in counts?? No matter what I do, I see the same counts. I am using "Self" Calibration. I have imported the mdb file into a "modern" 10.5.3 version and see the same behavior. Suggestions?

Probeman · « **Reply #11 on:** January 06, 2015, 11:04:36 AM »

Quote from: JohnF on January 06, 2015, 09:58:01 AM

I am trying to toggle the TDI on and off to see "what if". I am using the 9.6.3 PfE flavor on our SX51 which will continue operation for several more years. Regardless of _3_ ways of trying to turn off the TDI to see what the integrated count would be, nothing seems to work. The 3 ways are (1) in the Analytical/Analysis Options, uncheck the Use assigned or Self TDI corrections on Unknown box, (2) Assignment properties, click the No TDI Califbration Correction radio button, (3) under Standard and Interference Assignments, click the Remove TDI Correction. Have other people successfully been able to 'turn off' and see the difference in counts?? No matter what I do, I see the same counts. I am using "Self" Calibration. I have imported the mdb file into a "modern" 10.5.3 version and see the same behavior. Suggestions?

John,
The way to tell if the TDI corrections are on or off is to look for this output in the log window when the quant is calculated:

TDI%: 5.145 .038 ---- .222 ---- .434 -.063 ---- ---- ---- ---- ---- ---- DEV%: .4 .1 ---- .1 ---- .4 1.0 ---- ---- ---- ---- ---- ---- TDIF: LOG-LIN LOG-LIN ---- LOG-LIN ---- LOG-LIN LOG-LIN ---- ---- ---- ---- ---- ---- TDIT: 99.17 49.67 ---- 58.33 ---- 98.50 37.67 ---- ---- ---- ---- ---- ---- TDII: 254. 9873. ---- 3037. ---- 230. 146. ---- ---- ---- ---- ---- ----

You can also use the Remove TDI Correction button from the Standard Assignments dialog.

John Donovan · « **Reply #12 on:** March 20, 2015, 04:33:00 PM »

Quote from: Owen Neill on March 20, 2015, 02:43:13 PM

I'm assuming this error is a result of poor sequencing choices on my part, so obviously I need to structure my future setups so that no TDI's are needed for the last-measured sample of any "Combine Samples" projects. In this case, the elements for which TDI's were really necessary were all the second sample because... well, I'm an idiot. However, if anyone can suggest a workaround, I'd be mighty grateful.

Hi Owen,
Sad to say, the "loose nut behind the wheel" in this case is me!

I'll have to add code to handle the TDI intensities in the second sample. The issue is I need to properly re-normalize the TDI intensities from the 2nd condition to be consistent with the first sample conditions.

Please send me a small MDB file example of the issue and I will work on it ASAP.

Thanks!
john

Probeman · « **Reply #13 on:** March 30, 2015, 01:17:27 PM »

I took a shot this morning and acquired some time dependent intensity (TDI) data on the NBS K-373 and K-375 Na-Zn-Ba silicate glasses. Conditions were 15 keV, 30 nA and 5 um beam. 20 TDI and absorbed current measurements were made over a 30 sec counting time (using MAN bgds) as seen here:

and here:

Because of the extreme beam sensitivity of these two materials (10.4 elemental wt% Na), the double exponential (log-log), extrapolation to zero time is required. Quantification gives the following results for K-373:

ELEM: Na Si Zn Ba U O SUM 10 9.772 31.850 4.940 10.380 .054 42.340 99.336 11 9.944 31.850 4.940 10.380 .054 42.340 99.508 12 9.336 31.850 4.940 10.380 .054 42.340 98.900 AVER: 9.684 31.850 4.940 10.380 .054 42.340 99.248 SDEV: .314 .000 .000 .000 .000 .000 .314 SERR: .181 .000 .000 .000 .000 .000 %RSD: 3.24 .00 .00 .00 .00 .00 PUBL: 10.430 31.850 4.940 10.380 .054 42.340 99.994 %VAR: -7.15 .00 .00 .00 .00 .00 DIFF: -.746 .000 .000 .000 .000 .000 STDS: 301 0 0 0 0 0 STKF: .0510 .0000 .0000 .0000 .0000 .0000 STCT: 2902.3 .0 .0 .0 .0 .0 UNKF: .0470 .0000 .0000 .0000 .0000 .0000 UNCT: 2671.4 .0 .0 .0 .0 .0 UNBG: 27.7 .0 .0 .0 .0 .0 ZCOR: 2.0624 .0000 .0000 .0000 .0000 .0000 KRAW: .9204 .0000 .0000 .0000 .0000 .0000 PKBG: 97.30 .00 .00 .00 .00 .00 TDI%: 270.919 ---- ---- ---- ---- ---- DEV%: .6 ---- ---- ---- ---- ---- TDIF: LOG-LOG ---- ---- ---- ---- ---- TDIT: 53.67 ---- ---- ---- ---- ---- TDII: 2698. ---- ---- ---- ---- ----

and here for K-375:

ELEM: Na Si Zn Ba U O SUM 13 9.460 31.830 4.940 10.370 .110 42.320 99.030 14 8.583 31.830 4.940 10.370 .110 42.320 98.153 15 9.695 31.830 4.940 10.370 .110 42.320 99.265 AVER: 9.246 31.830 4.940 10.370 .110 42.320 98.816 SDEV: .586 .000 .000 .000 .000 .000 .586 SERR: .339 .000 .000 .000 .000 .000 %RSD: 6.34 .00 .00 .00 .00 .00 PUBL: 10.420 31.830 4.940 10.370 .110 42.320 99.990 %VAR: -11.27 .00 .00 .00 .00 .00 DIFF: -1.174 .000 .000 .000 .000 .000 STDS: 301 0 0 0 0 0 STKF: .0510 .0000 .0000 .0000 .0000 .0000 STCT: 2902.3 .0 .0 .0 .0 .0 UNKF: .0447 .0000 .0000 .0000 .0000 .0000 UNCT: 2543.9 .0 .0 .0 .0 .0 UNBG: 27.8 .0 .0 .0 .0 .0 ZCOR: 2.0680 .0000 .0000 .0000 .0000 .0000 KRAW: .8765 .0000 .0000 .0000 .0000 .0000 PKBG: 92.61 .00 .00 .00 .00 .00 TDI%: 236.546 ---- ---- ---- ---- ---- DEV%: .6 ---- ---- ---- ---- ---- TDIF: LOG-LOG ---- ---- ---- ---- ---- TDIT: 52.67 ---- ---- ---- ---- ---- TDII: 2568. ---- ---- ---- ---- ----

Although the relative accuracy errors are still around 10%, the TDI correction is over 200%! What does that mean? Well, it means that if you didn't use the TDI correction, you'd get a roughly 70% error, even with these relatively mild conditions:

ELEM: Na Si Zn Ba U O SUM 13 2.929 31.830 4.940 10.370 .110 42.320 92.499 14 2.866 31.830 4.940 10.370 .110 42.320 92.436 15 2.852 31.830 4.940 10.370 .110 42.320 92.422 AVER: 2.882 31.830 4.940 10.370 .110 42.320 92.452 SDEV: .041 .000 .000 .000 .000 .000 .041 SERR: .024 .000 .000 .000 .000 .000 %RSD: 1.41 .00 .00 .00 .00 .00 PUBL: 10.420 31.830 4.940 10.370 .110 42.320 99.990 %VAR: -72.34 .00 .00 .00 .00 .00 DIFF: -7.538 .000 .000 .000 .000 .000 STDS: 301 0 0 0 0 0

But I'm also curious about the fact that the two very similar materials show changes in absorbed current that are opposite of each other even though they have just a slight difference in uranium content.

Probeman · « **Reply #14 on:** March 30, 2015, 02:23:19 PM »

Quote from: Probeman on March 30, 2015, 01:17:27 PM

Although the relative accuracy errors are still around 10%, the TDI correction is over 200%! What does that mean? Well, it means that if you didn't use the TDI correction, you'd get a roughly 70% error, even with these relatively mild conditions...

Here are the same data but with the "TDI weighting" feature applied as seen here:

Here are the results for the K-373 standard. Note that we now obtain around 5% relative accuracy even with an almost 300% TDI correction!

ELEM: Na Si Zn Ba U O SUM 10 9.992 31.850 4.940 10.380 .054 42.340 99.556 11 10.468 31.850 4.940 10.380 .054 42.340 100.032 12 9.130 31.850 4.940 10.380 .054 42.340 98.694 AVER: 9.864 31.850 4.940 10.380 .054 42.340 99.428 SDEV: .678 .000 .000 .000 .000 .000 .678 SERR: .392 .000 .000 .000 .000 .000 %RSD: 6.88 .00 .00 .00 .00 .00 PUBL: 10.430 31.850 4.940 10.380 .054 42.340 99.994 %VAR: -5.43 .00 .00 .00 .00 .00 DIFF: -.566 .000 .000 .000 .000 .000 STDS: 301 0 0 0 0 0 STKF: .0510 .0000 .0000 .0000 .0000 .0000 STCT: 2939.1 .0 .0 .0 .0 .0 UNKF: .0479 .0000 .0000 .0000 .0000 .0000 UNCT: 2758.6 .0 .0 .0 .0 .0 UNBG: 27.8 .0 .0 .0 .0 .0 ZCOR: 2.0602 .0000 .0000 .0000 .0000 .0000 KRAW: .9386 .0000 .0000 .0000 .0000 .0000 PKBG: 100.40 .00 .00 .00 .00 .00 TDI%: 283.286 ---- ---- ---- ---- ---- DEV%: .5 ---- ---- ---- ---- ---- TDIF: LOG-LOG ---- ---- ---- ---- ---- TDIT: 53.67 ---- ---- ---- ---- ---- TDII: 2781. ---- ---- ---- ---- ----

This is what it looks like with the weighted fitting. Still slightly, slightly under fit...

Note that this is a log-log (double exponential) fit and still not quite enough curvature. Now that's what I call Na loss! But as long as the precision of our measurements are good enough, we can continue to push our modeling accuracy.

News:

Author Topic: Time Dependent Intensity (TDI) Corrections (Read 29541 times)

John Donovan

Time Dependent Intensity (TDI) Corrections

John Donovan

Time Dependent Intensity (TDI) Corrections

John Donovan

Re: Time Dependent Intensity (TDI) Corrections

John Donovan

Re: Time Dependent Intensity (TDI) Corrections

Richard Walshaw

Re: Time Dependent Intensity (TDI) Corrections

Probeman

Re: Time Dependent Intensity (TDI) Corrections

Probeman

Re: Time Dependent Intensity (TDI) Corrections

John Donovan

Re: Time Dependent Intensity (TDI) Corrections

Gareth D Hatton

Re: Time Dependent Intensity (TDI) Corrections

Probeman

Re: Time Dependent Intensity (TDI) Corrections

JohnF

Re: Problem with Time Dependent Intensity (TDI) Corrections: turning it off.

Probeman

Re: Problem with Time Dependent Intensity (TDI) Corrections: turning it off.

John Donovan

Re: Time Dependent Intensity (TDI) Corrections

Probeman

Re: Time Dependent Intensity (TDI) Corrections (Absorbed current)

Probeman

Re: Time Dependent Intensity (TDI) Corrections (Absorbed current)