Performing Integrated WDS and EDS Acquisition in PFE

[John Donovan]

While Ryan McAleer and I were looking into how we might decrease the time it takes to start an EDS acquisition using the JEOL MEC EDS interface, when using the TDI acquisition feature in Probe for EPMA:

https://smf.probesoftware.com/index.php?topic=79.msg13791#msg13791

we found another "interesting" behavior of the JEOL EDS software.

Previously we found that the JEOL EDS software automatically inserts the Faraday cup (PCD) when the EDS acquisition is complete. Of course when one is still acquiring WDS elements, this is not a good thing.  So in addition to item #2 listed here:

Ryan McAleer makes some suggestions for getting the best results from the JEOL EDS interface on iSP100/iHP200F instruments:

1. You should run a single EDS spectrum acquisition in the JEOL software prior to any EDS acquisitions with Probe for EPMA. Doing seems to get the JEOL software interface into a correct state. Just one EDS acquisition for each time starting the JEOL software. 

2. To avoid issues with the Faraday cup inserting before the WDS acquisitions are finished, you must uncheck the "Insert PCD after reservation execution is completed" [PFE will prompt you about this once per run]

3. The user should run a single EDS acquisition after they change the EDS time constant in Probe for EPMA. This is to ensure that these EDS conditions are properly saved for subsequent EDS element analyses. Or one can just set the EDS time constant they want in both JEOL (Analysis-->EDS analysis basic conditions) and Probe for EPMA. Otherwise the documented time constant on that first point will be incorrectly saved [we are looking into why this occurs].

4. If there is any hangup during EDS comm, RESTART the JEOL COMPUTER.  If you just restart the software (either through "ResetEPMA" or through the task manager) the "reservation" of the EDS will not clear and you will run into a variety of problems.

we have to add a new item because the JEOL EDS software also seems to change the beam mode once the EDS acquisition is completed, even if WDS elements are still being acquired!

Here are Ryan's observations:

During EDS analysis: JEOL software indicates scan coils on and spot mode--equivalent of pfe digital spot mode I think.



After EDS ends (but WDS still going), changes to this configuration: pfe equivalent of scan mode. I think when EDS run ends JEOL reverts to a standard state, which is probe scan, note that the beam stays defocused if you have that set, so if the Mag is set to 270K for analysis then maybe there is little practical effect. Becomes an issue if you want to make a valiant attempt to measure something smaller than FOV at 270K (<500nm).

So, just FYI.

[Ben Buse]

Hi a question about pulling intensities...

The PFE software stores the EDS spectra at time of analysis, and then at a later time it asks the JEOL EDS to extract the net intensity from the spectra stored in PFE - is that correct?

Which leads to two further questions

a. Can the PFE ask the JEOL offline software PC-SEM to do this task?

b. If PFE had the spectra for each pixel in a map (if this was extracted from JEOL pts and given to PFE), could it then ask the JEOL EDS to give net intensities for each spectra?

 

[Ben Buse]

With rgds to question a. interestingly testEDS works, on JEOL offline computer PC-SEM

But PFE doesn't seem to

[John Donovan]

With rgds to question a. interestingly testEDS works, on JEOL offline computer PC-SEM,
But PFE doesn't seem to

That's probably because you are trying to run PFE at the same time TestEDS is connected and the JEOL LibEDSIS DLL won't allow that because it isn't multi-threaded.

The PFE software stores the EDS spectra at time of analysis, and then at a later time it asks the JEOL EDS to extract the net intensity from the spectra stored in PFE - is that correct?

Not quite.

PFE acquires the EDS point spectra at time of acquisition and stores it to the run database. At the same time, PFE extracts the net intensities and stores them also, unless during an automated acquisition with EDS the user answered the question "Do you want to store EDS net intensities during the automation?" with a No, as described here:

https://smf.probesoftware.com/index.php?topic=79.msg13471#msg13471

The reason for this question is because when one has many points per sample, every time PFE acquires a new data point, it lists the net intensities for both WDS and EDS elements for that sample.  For WDS elements there is essentially no "overhead" because the WDS data is stored in the local MDB file.  But for EDS elements PFE needs to query the EDS server and obtain the net intensities from the OEM spectrum deconvolution software, which takes a few seconds per point. 

If there are hundreds of points per sample, this starts to add up, so the user is asked whether they want to do this net intensity extraction for every point acquisition, or do it later after the automation has completed, when it will take much less time (because it only needs to obtain the net intensities once per sample/point).

So yes, normally PFE acquires the EDS spectrum, obtains the EDS net intensities for any EDS elements, and then stores those also. If EDS elements are added subsequently to the sample, when the quant is run again, or even just the raw data is displayed, PFE will again obtain the EDS net intensities and store them again.

Which leads to two further questions

a. Can the PFE ask the JEOL offline software PC-SEM to do this task?

Yes, but there's a catch for the JEOL software.

At acquisition time, the EDS spectrum for each point analysis is stored in the local PFE database obtained from the JEOL (and Thermo and Bruker) server software. At that time (unless turned off by the user during automation as described above), PFE also obtains the net intensities for each EDS element and also stores those net intensities locally.

For the Thermo and Bruker EDS software, this means reading the EDS spectrum stored locally in the PFE database, and sending that spectrum, along with the spectrum meta data, EDS elements/x-ray lines, to the OEM EDS server, and then getting the EDS net intensities back, which are then (as previously mentioned) also stored locally.

For the JEOL software however, the JEOL EDS server cannot receive an EDS spectrum and meta data and return the EDS net intensities. Instead, the JEOL EDS server needs the *name* (GUID actually) of the EDS point acquisition which is stored locally on the JEOL server. 

When this GUID is passed to the JEOL EDS server, the JEOL server software reads the EDS spectrum and meta data from its local storage and returns the net intensities for the EDS elements specified by PFE.

Why this matters for obtaining EDS net intensities, is because for the Thermo and Bruker EDS systems, one can simply install the Thermo of Bruker EDS software on any computer and when running Probe for EPMA off-line for quantitative analyses, one can simply send the EDS spectrum (and meta data) to the OEM EDS software and the OEM software will return the EDS net intensities to PFE for quantitative analyses. Even when PFE or the computer is not connected to the instrument!

However, with the JEOL EDS software, even if one installed the JEOL software on an off-line computer, one could still not obtain the EDS net intensities, because the EDS spectrum files are stored with the JEOL EDS software which is on the JEOL computer connected to the instrument!

However, there is a "work around" for this. If one wants to re-process EDS elements off-line on another computer, that is not connected to the OEM software on the OEM computer connected to the instrument, one can simply make sure that all net intensities are extracted and stored in the PFE database locally by just having the PFE software list the raw data for each sample from the Analyze! window. Then go into the Analytical | Analysis Options menu and check this checkbox:



Then PFE will only use locally stored EDS net intensities for quantitative analyses and this file can be moved to any computer for further off-line re-processing.

b. If PFE had the spectra for each pixel in a map (if this was extracted from JEOL pts and given to PFE), could it then ask the JEOL EDS to give net intensities for each spectra?

It's a great question with a very complicated answer.

The EDS API software system on your instrument did have functions for this (LibEDSIS.DLL), e.g., EDSGetNetCountMapData(), but by the time we were implementing this feature, JEOL stopped supporting this 8230/8530 EDS API, and now only support the new MEC EDS API (for the iSP100/iHP200F), which so far as I can tell does not have support for EDS net intensities from spectrum images.  I can ask JEOL Japan again, but it wouldn't help you on the 8230/8530 instrument with the old EDS API.

I can't find the email, but apparently there could also be an issue with getting the JEOL EDS hypermap net intensities when the spectrum image data is acquired from the WDS scan generator vs. the JEOL video scan generator.  I remember they mentioned that getting the EDS pixels hardware dead time corrected in the WDS hypermap was "very difficult". Apparently only the EDS hypermaps acquired using the video scan generator are corrected for dead time using hardware. Which is why we started trying to get Bruker to work with JEOL to do synchronized WDS and EDS mapping.

If you want to contact me directly by email or Zoom we can discuss what we might be able to do for your situation.

[John Donovan]

As we have mentioned previously the JEOL PC-SEM software automatically inserts the PCD (Faraday cup) after the EDS spectrum acquisition is complete (it's a feature not a bug!).   

In order to prevent the Faraday cup from being inserted before the WDS acquisition finishes (if it is using a longer acquisition time than the EDS), one must be sure to *uncheck* the "Insert PCD after the reservation execution is completed" checkbox in the JEOL software:



A slightly more subtle bug (sorry, I mean feature!), is that the JEOL EDS software also changes the beam mode from point to scanning mode once the EDS acquisition is complete.  So if you haven't specified a high magnification for scanning, say 100Kx and your WDS takes longer than the EDS acquisition, you might have an issue when your WDS quantification when the EDS completes.

I asked JEOL Japan if they could add another check box to uncheck so the JEOL EDS software would not change the beam mode when the EDS acquisition is complete, but they said they looked into it and it would be too difficult do. I have to wonder if this same thing happens with the JEOL PC-EPMA software when acquiring WDS and EDS together?  Does anyone know?

Anyway, just a heads up so you know when performing integrated WDS and EDS in Probe for EPMA using the JEOL EDS that you make sure your EDS acquisition times are always longer than your WDS acquisition times.

[JonF]

I asked JEOL Japan if they could add another check box to uncheck so the JEOL EDS software would not change the beam mode when the EDS acquisition is complete, but they said they looked into it and it would be too difficult do. I have to wonder if this same thing happens with the JEOL PC-EPMA software when acquiring WDS and EDS together?  Does anyone know?

I tried combined WDS/EDS on our iHP200F using SEM Center and PC-EPMA, using all three options of "WDS -> EDS", "EDS -> WDS" and "WDS/EDS" in the PC-EPMA software. I set a single WDS pass of elements for 20s on peak plus 5s off peak and 10s EDS real time, and in each case it didn't change the beam conditions until after the acquisition was completed for all spectrometers (EDS and WDS).     
This is assuming the SEM Center GUI is updated whilst the analysis is ongoing, but to try and see if this is the case, I set up test points on a cathodoluminescent sample and it didn't seem to change until the acquisition is finished.

So it would seem that PC-EPMA can do what you suggested?

[John Donovan]

I asked JEOL Japan if they could add another check box to uncheck so the JEOL EDS software would not change the beam mode when the EDS acquisition is complete, but they said they looked into it and it would be too difficult do. I have to wonder if this same thing happens with the JEOL PC-EPMA software when acquiring WDS and EDS together?  Does anyone know?

I tried combined WDS/EDS on our iHP200F using SEM Center and PC-EPMA, using all three options of "WDS -> EDS", "EDS -> WDS" and "WDS/EDS" in the PC-EPMA software. I set a single WDS pass of elements for 20s on peak plus 5s off peak and 10s EDS real time, and in each case it didn't change the beam conditions until after the acquisition was completed for all spectrometers (EDS and WDS).     
This is assuming the SEM Center GUI is updated whilst the analysis is ongoing, but to try and see if this is the case, I set up test points on a cathodoluminescent sample and it didn't seem to change until the acquisition is finished.

So it would seem that PC-EPMA can do what you suggested?

Interesting.  Thanks for testing this.

It seems that JEOL MEC EDS interface API is based on the JEOL PC-SEM application GUI rather than the actual sub routines available to the PC-EPMA softare. Too bad... maybe you should have bought a Bruker!   

[Ben Buse]

Hi

Is it possible to add EDS elements after the analysis has been completed if you've acquired the spectrum?

[John Donovan]

Is it possible to add EDS elements after the analysis has been completed if you've acquired the spectrum?

Of course you can.

You can add an element by EDS before or after the sample has EDS spectrum data. Here is an example of adding an element by EDS for use in the interference correction with another element (WDS or EDS) that is already acquired:

https://smf.probesoftware.com/index.php?topic=482.msg2826#msg2826

This hasn't been tested with WDS and EDS maps, but it should be the same in CalcImage (if we ever get that working):

https://smf.probesoftware.com/index.php?topic=1524.msg11777#msg11777