JEOL SXES Soft X-ray Emission Spectrometer

[Dan MacDonald]

Good morning, everyone:

Has anyone out there ever bought or used the JEOL SXES soft x-ray detector, especially for Li determination?  Likes or dislikes? Pros and cons?  Ease of use vs difficulties and limitations?  We are looking at purchasing one of these systems for our JXA 8200, and I thought I'd ask, in case anyone outthere had any helpful advice.  Thanks very much!

Kind regards

Dan MacDonald
Probe Dude
Dalhousie University

[Probeman]

Has anyone out there ever bought or used the JEOL SXES soft x-ray detector, especially for Li determination?  Likes or dislikes? Pros and cons?  Ease of use vs difficulties and limitations?  We are looking at purchasing one of these systems for our JXA 8200, and I thought I'd ask, in case anyone outthere had any helpful advice. 

Hi Dan,
I didn't know that JEOL was making this spectrometer available for older instruments...  very cool.

I wonder if JEOL would be willing to provide an API (application programming interface) so one could acquire these low energy spectra into Probe for EPMA.  Similar to how we acquire already spectra for EDS and CL in Probe for EPMA. You should be sure to get a commitment from JEOL for them to provide such an API as part of your SXES spectrometer purchase if you decide to do this.

I not sure what exactly one could do with these spectra with regards to quantification (calibration curve methods?), but getting them into a database would be a great start.

I suppose this interest is because Dalhousie is a center for Li- ion battery research? 

http://www.dal.ca/diff/dahn/people/jeff_dahn.html

Pretty exciting stuff.
john

[Anette von der Handt]

I have some experience. I send you a message.

[John Donovan]

I have some experience. I send you a message.

Please send me a message as well!

[jon_wade]

and here please....

[DirkMueller]

Hi all,

I found one discussion about this JEOL detector from the year 2016, which was only some posts long - therefore, I decided to start a new one.

a) I would be interested if someone has experience / recommendations for using the SXES-ER detector as a substitution for one WDS spec (since the SXES would require one WDS port)?
--> The energy range of the JS2000 model (350-2300 eV) would cover the Ka energies for elements from N to P. The energy resolution would be comparable to a WDS system.

b) Has someone experience to determine Fe2+/Fe3+ (or valence states for other elements) by using the SXES detector?

Thanks
Dirk

[Mike Matthews]

I believe there’s now an option to fit it to an EDS port rather than sacrifice a WDS.

[Gareth D Hatton]

We have the SXES-ER on our EDS port and an EDS on the other port so we have not sacrificed a WDS.

[John Donovan]

I found one discussion about this JEOL detector from the year 2016, which was only some posts long - therefore, I decided to start a new one.

We merged these two topics as it is probably better to keep these posts together.

As you mentioned there are some other posts discussing SXES spectrometers. See also these posts where Jon Fellowes asks for some mods to the PFE software (dead time sanity check) to allow him to treat the SXES spectrometer as a WDS spectrometer.:

https://probesoftware.com/smf/index.php?topic=71.msg11872#msg11872

[DirkMueller]

Thanks for the info regarding the port occupation (from Jeol I got the info that one WDS port is required  ) - I'll ask one more time...

However, my question was more, if someone uses the SXES detector for full-quantitative analysis? If it is comparable to a WDS?

So far, I can only see benefits:
- faster analysis, since all elements are measured simultaneously
- no need for regular calibration, since no moving parts and no gas (maybe temperature plays a role?)

[JonF]

I'll chime in here, as I get a lot of use out of our SXES (-LR rather than -ER though).

Regarding the WDS vs EDS port mounting, it depends on what chamber you have. The 8x30 series chambers (8230 and 8530F) and earlier have the SXES fitted to either WDS port 1 or 5 (i.e the two ports either side of the airlock), whereas I've been told the new iHP200F has a modified EDS port so that you can fit the SXES there instead (with a higher take off angle). You can see this arrangement on the front cover of the JEOL SXES brochure (from https://www.jeol.com/download_catalogues.php#sem_fe_op).

If you're looking for a new machine, then you'll be asking for the SXES-EREP (or the SXES-LREP), where the EP = "EDS Port".

I've spent quite a lot of time getting my head around the best way of using the SXES-LR, as its quite a different beast compared to the WDS and EDS. The -LR (low range) comes with the JS50XL and the JS200N diffraction gratings, and these really are low energy (the energy of the C Ka is way too high for example, I have to use the 2nd order reflection or higher). The JS300N grating in the SXES-ER(/EP) looks to be pretty similar to the JS200N, so it'll be really for light element/soft X-ray use (and the all the complications that entails). 
The JS2000 has a waaaay bigger energy range than the other three diffraction crystals, ranging from the smaller d spacing LDE crystals up through the TAP range and in to the bottom of the PET energy range. The CSIRO folks had a talk a while back where they compared the spectral resolution of the JS2000 grating across its range and found it was favourable relative to TAP for most of that range, but is notably worse than PET at the high energy end. 

The big draw back of the SXES that I can think of are:
(i) time - and lots of it! - the SXES CCD is pretty slow. This is negated slightly by counting all elements and all backgrounds simultaneously, but you'll still be counting for 100s if not 1000s of msec per spectrum at the very least. This makes for some very long maps. Quantitative points would be possible, but you'll have some software to write!
(ii) it's bloomin' cold. The SXES CCD sits unprotected in your sample chamber at -70C, ensuring that any contamination you have in your system will sit on the CCD pretty rapidly - you're effectvely cryo-pumping your sample chamber. You'll either need to use the liquid nitrogen anticontamination device permanently (which is a nuisance when you have to fill up the dewar twice a day, and letting it run dry will crash the vacuum), or design your own anti-contamination device (I've built an in vacuo Peltier based system).

Overall, I think the SXES is worth it. It's a big step forward in light element analysis, with much higher spectral resolution and essentially no deadtime to worry about. I don't think it's a replacement for WDS and GFPC, but it's another useful detector alongside the WDS and EDS.

[John Donovan]

The big draw back of the SXES that I can think of are:
(i) time - and lots of it! - the SXES CCD is pretty slow. This is negated slightly by counting all elements and all backgrounds simultaneously, but you'll still be counting for 100s if not 1000s of msec per spectrum at the very least. This makes for some very long maps. Quantitative points would be possible, but you'll have some software to write!
...
Overall, I think the SXES is worth it. It's a big step forward in light element analysis, with much higher spectral resolution and essentially no deadtime to worry about. I don't think it's a replacement for WDS and GFPC, but it's another useful detector alongside the WDS and EDS.

Awesome information, thanks for sharing this!

Previously in this topic you asked us (Probe Software) to modify the allowable range of dead time values (and we did) because you said you wanted to reprocess SXES data using our software. Could you expand upon that aspect?  How exactly do you utilize SXES data in our software?   You acquire the SXES data with your own app? Then you write it yourself as a PrbImg file? And then reprocess it in CalcImage?  Could you detail these steps for us?

We are very curious how you do these things and if there's anything we could do to further facilitate this processing.

[JonF]

Awesome information, thanks for sharing this!

Previously in this topic you asked us (Probe Software) to modify the allowable range of dead time values (and we did) because you said you wanted to reprocess SXES data using our software. Could you expand upon that aspect?  How exactly do you utilize SXES data in our software?   You acquire the SXES data with your own app? Then you write it yourself as a PrbImg file? And then reprocess it in CalcImage?  Could you detail these steps for us?

We are very curious how you do these things and if there's anything we could do to further facilitate this processing.

Actually, I eventually decided against reinventing the wheel and just went with acquiring all the signals with the JEOL software (SE, BSE, CL, EDS, WDS and SXES).
Then its just the comparatively simple job of converting between file formats (SXES -> GRD) to get the information in to CalcImage, akin to the file conversions available in Probe Image.

[John Donovan]

Previously in this topic you asked us (Probe Software) to modify the allowable range of dead time values (and we did) because you said you wanted to reprocess SXES data using our software. Could you expand upon that aspect?  How exactly do you utilize SXES data in our software?   You acquire the SXES data with your own app? Then you write it yourself as a PrbImg file? And then reprocess it in CalcImage?  Could you detail these steps for us?

We are very curious how you do these things and if there's anything we could do to further facilitate this processing.

Actually, I eventually decided against reinventing the wheel and just went with acquiring all the signals with the JEOL software (SE, BSE, CL, EDS, WDS and SXES).
Then its just the comparatively simple job of converting between file formats (SXES -> GRD) to get the information in to CalcImage, akin to the file conversions available in Probe Image.

That sounds quite reasonable. 

Have you attempted to quantify any SXES signals in CalcImage?

[JonF]

That sounds quite reasonable. 

Have you attempted to quantify any SXES signals in CalcImage?

Yeah, it works really well! I'll speak with a couple of recent users and see if I can get some examples up.