Author Topic: Long Arbitrarily Oriented Traverses  (Read 37 times)

glennpoirier

  • Professor
  • ****
  • Posts: 44
Long Arbitrarily Oriented Traverses
« on: October 16, 2020, 01:23:40 pm »
Hi All,
I'm trying to acquire some long closely-spaced traverses across growth zones in Mn nodules. I've done this before using very thin quant maps, but unless I'm missing something, you can't do strip maps that are at an arbitrary angle to the X or Y axes. I was thinking that I could just do parallel long traverses, but there's a problem there too. I want to use really short counting time (ca. 1s) so the spectrometer move times will add a significant multiplier to the analysis time. Ideally I'd like to measure the first five elements over the whole traverse and then come back and repeat the points for the second 5 elements. Is this possible in PfE? I couldn't find a switch for this.
What I think I'm going to have to do is make two different sample setups  and run the traverse once for each setup. Then I can use the Combine Selected Samples button to combine them for the matrix correction.
Can anyone verify that this is the easiest way to do this, or am I missing something?

Have a great weekend
Glenn

Probeman

  • Emeritus
  • *****
  • Posts: 2079
  • Never sleeps...
    • John Donovan
Re: Long Arbitrarily Oriented Traverses
« Reply #1 on: October 16, 2020, 05:07:06 pm »
Hi All,
I'm trying to acquire some long closely-spaced traverses across growth zones in Mn nodules. I've done this before using very thin quant maps, but unless I'm missing something, you can't do strip maps that are at an arbitrary angle to the X or Y axes. I was thinking that I could just do parallel long traverses, but there's a problem there too. I want to use really short counting time (ca. 1s) so the spectrometer move times will add a significant multiplier to the analysis time. Ideally I'd like to measure the first five elements over the whole traverse and then come back and repeat the points for the second 5 elements. Is this possible in PfE? I couldn't find a switch for this.
What I think I'm going to have to do is make two different sample setups  and run the traverse once for each setup. Then I can use the Combine Selected Samples button to combine them for the matrix correction.
Can anyone verify that this is the easiest way to do this, or am I missing something?

Have a great weekend
Glenn

Hi Glenn,
There are at least a couple of possible ways to proceed here.

1. You could certainly set up two separate point traverse setups in Probe for EPMA each with 5 elements (to save spectrometer motion time) and then, as you mentioned, utilize the Combine Selected Samples button to obtain a complete analysis.  That might be how I proceed.

2. Aside from rotating the sample to an orthogonal orientation in the stage holder, maybe JEOL has a map acquisition method to acquire a pixel mapping scan line at an arbitrary angle (though Probe Image does not), so then you could acquire your scan line that way and then to quantify it, utilize the conversion utility in Probe Image to convert the JEOL scan line map into the PrbImge format for subsequent quantification in CalcImage.

Alternatively you could specify a polygon mapped area using the JEOL software (which I know works), that delineates a diagonal strip across your Mn nodule.  And then, once again, utilize the convert utility in Probe Image to convert your JEOL map to a PrbImg file for subsequent quantification in CalcImage.

Finally it should be mentioned that if you do obtain a single pixel wide diagonal scan line map using the JEOL (or Cameca software) and want to subsequently quantify that in CalcImage, that is fine, but you will not be able to output the data using Golden Software's Surfer app as Surfer GRD files must be at least 2 pixels wide. For (orthogonal) single scanline maps, you would need to convert them to a 2 pixel wide scan as described here:

https://probesoftware.com/smf/index.php?topic=1154.0

I'm sure I've completely overlooked other methods, but that is my first thoughts on the question...
« Last Edit: October 16, 2020, 05:15:13 pm by Probeman »
The only stupid question is the one not asked!

glennpoirier

  • Professor
  • ****
  • Posts: 44
Re: Long Arbitrarily Oriented Traverses
« Reply #2 on: October 19, 2020, 12:47:23 pm »
Hi John,
Thanks for the detailed response. I think in the end I'll use one of JEOL's cup holder and fix the sample to the top. It will allow fairly simple rotation but it will  need a bit of futzing to get the orientation right (I'll probably get the shop to make me something easier to adjust).
A new question arises though.  The sample top is at Z = 12.65 and ProbeImage doesn't like this number. Where is probe image getting its stage limit numbers?

Cheers
Glenn

John Donovan

  • Administrator
  • Emeritus
  • *****
  • Posts: 2631
  • Other duties as assigned...
    • Probe Software
Re: Long Arbitrarily Oriented Traverses
« Reply #3 on: October 19, 2020, 01:29:21 pm »
A new question arises though.  The sample top is at Z = 12.65 and ProbeImage doesn't like this number. Where is probe image getting its stage limit numbers?

In Probe Image click the Setup | Probe Image Setup menu, Stage/Spectro Extents button.
John J. Donovan, Pres. 
(541) 343-3400

"Not Absolutely Certain, Yet Reliable"

glennpoirier

  • Professor
  • ****
  • Posts: 44
Re: Long Arbitrarily Oriented Traverses
« Reply #4 on: October 19, 2020, 01:29:42 pm »
Nevermind, found it. (Setup->  Probe Image Setup > Instrument > Stage/Spectro) Was looking for a config file.
Cheers
Glenn

Anette von der Handt

  • Global Moderator
  • Professor
  • *****
  • Posts: 263
    • UMN Probelab
Re: Long Arbitrarily Oriented Traverses
« Reply #5 on: October 19, 2020, 09:22:14 pm »
Hi Glenn,

a note of caution with the 1 second count time. In my experience there is sometimes a lag with the detector recording the X-ray counts in the first second. I had that on my 8900 and now on the 8530FPlus and I know of at least one other JEOL instrument. It really only comes out in 1 second measurements but maybe you have seen it in TDI measurements where occasionally the first point is 50% lower than the following counts.

You can often check it also by running the chart recorder, open the PCD and see how long it takes for the different spectrometers to register the counts (in my experience the time lag also differs between the spectrometers).

One way around is to set a 1 second incubation delay for the analyses if you find a similar behavior on your instrument.
Against the dark, a tall white fountain played.