Spectrometer configuration

[Jeremy Wykes]

I would like to canvas the EPMA community for their thoughts about spectrometer/crystal configuration.

What is your current spectrometer/crystal configuration, would you change it if you had the chance, and what sort of samples do you mainly analyse.

On the RSES SX100 we have:

SP1 TAP, PC0
SP2 PET, PC2
SP3 LPET, LLIF
SP4 LPET, LLIF

I find this configuration suboptimal. A single TAP crystal is inconvenient, given that many of our users are experimental petrologists analysing glasses.

[Anette von der Handt]

Hi, University of Minnesota chiming in.

Our current layout is (JEOL-JXA8900):

SP1: LDE1, TAP, PETJ
SP2: LDE1, PETJ, TAP, LDEB
SP3: LDE2, TAP
SP4: PETJ, LIF,
SP5: PETH, LIFH

Overall, I am very happy with it. I think it was put together by Pete McSwiggen and I haven't run into many situations where I would have wanted a different layout. We have similar mixed-bag applications to what Owen reported except that no one has bothered me with Boron yet (but we do C, N, O work heavily).

However, I concur with Owen, if/when I get a new probe and have the appropriate funds I would opt for more L-type channels as we do a lot of trace element work and look more into what innovations are out there for our light element work.

[Karsten Goemann]

Uni of Tasmania SX100, #846, installed 2003, configured by my predecessor David Steele:

1: TAP, PET, PC0, PC2, low pressure
2: PET, LLiF, high pressure
3: LPET, LiF220, high pressure
4: TAP, PC1, low pressure
5: LPET, LLiF, high pressure

We do a lot of sulphide minerals, REE minerals... so the "heavy" configuration suits us.
I don't think I could go back to 4 WDS and not being able to do anything between 2TAP+3PET or 1TAP+2LiF+2PET. Even now it would be great to be able to do combined WDS-EDS (proper EDS = no ROIs) and do some of the majors by EDS if precision is similar.

The LiF220 was an experiment for higher energy lines, but because it is not a large area (high intensity) crystal and a high percentage of the x-rays basically go straight through the P10 the count rates are not good, so I probably wouldn't do it again. Just put another LLiF on there.

But most days we have SP3 on LPET, SP4 on TAP, and SP5 on LLiF and just change crystals on SP1 and 2 depending on the element list.

[Gseward]

UCSB SX100: #921 installed August 2010.

1) LTAP LPET; Low Pressure P10
2) PC0 PC2 TAP PET; Low Pressure P10
3) LPET LLIF; High Pressure P10
4) LTAP LPET; Low Pressure P10
5) LPET LLIF; High Pressure P10
6) Thermo Ultra Dry EDS.

Mixed bag of samples come through the lab, from Geological materials to experimental aerospace alloys and jet engine thermal barrier coatings. Trace element work was the focus when the instrument was purchased. Mostly I'm happy with the crystal config (I have to be since I picked it!). Some days I'd like 3 LLIFs, for REE work, but then those days I'd also like 15 spectrometers. 

[Karsten Goemann]

Regarding the LiF220, yes U and Th L are available, and also K lines for some interesting elements. If I've done the calculations correctly:

Line - keV - sine theta
NbKa1 - 16.6 - 0.262
ZrKa1 - 15.78 - 0.276
YKa1 - 14.96 - 0.291
SrKa1 - 14.16 - 0.307
...
ULa1 - 13.61 - 0.320
ThLa1 - 12.97 - 0.336
...

On the Cameca you can get ThLa1 on a "normal" (L)LiF but it is close to the limit (sine theta 0.237, lower spectro limit usually something like 0.224), ULa1 is at 0.226, so pretty much at the limit.

But if we analyse those elements they're usually in small, inhomogenous grains of accessory minerals in rocks, so I tend to try to avoid having those x-rays where I can (by keeping the accelerating voltage below the LIII edge of U and Th, for example), as these will go for miles in the material.

Our probe also doesn't really like to be at accelerating voltages above 25kV.

Cheers,

Karsten

[Probeman]

Here's the config on my SX100 at UofO (purchased in 2005):

Spec 1, (PET, TAP, PC1, PC2), low pressure
Spec 2, (LPET, LTAP), low pressure
Spec 2, (LLIF, LPET), high pressure
Spec 4 (PET, TAP, PC1, PC25), low pressure
Spec 5, (LIF, PET), high pressure

Some notes:

1. This instrument was going to be dry pumped but the Euro did not cooperate so I got a diff pump and a 100 kelvin chilled baffle that has been absolutely trouble free and we have extremely low carbon contamination rates similar to cold finger and oxygen jet systems without the hassle. See :

http://probesoftware.com/smf/index.php?topic=140.msg566#msg566

2. The two 4 crystal spectros are necessary because we do a lot of B, C, N and O analyses. In fact the PC25 is optmizied for B ka.

3. If I had to do it all over again I would get large area crystals on spec 5.

4. I'd also insist on a pin diode detector on the exit window of the high pressure detectors as discussed here to significantly improve high energy sensitivity:

http://probesoftware.com/smf/index.php?topic=193.msg1831#msg1831

I'll also add that having PET crystals on all spectrometers is useful for several operations, e.g., dead time calibrations and simultaneous k-ratios checks using Si ka (PET/TAP) and Ti Ka (PET/LIF), sulfur chemical shift measurements, and Ti in quartz on all 5 spectrometers...

[Gareth D Hatton]

Here is the configuration of our JXA 8500F

Spec 1, 4 crystal Gas flow (PET, TAP, LDE1, LDE2)
Spec 2, Gas flow (PET, TAP)
Spec 3, Xenon H-type (PETH, LiFH)
Spec 4, Gas flow (PET, TAP)
Spec 5, 4 Xenon (PET, LiF)

Lots of PGE analysis hence the PET crystals.  Some days this is an ideal configuration but there are other days when the request is for too many transition elements that are 'best' on LiF  .  We also have an integrated EDS and cold finger.  If I could change anything I think I would have another EPMA as well Perhaps with the new SXES, EDS and all the toys needed for a clean system...

[joeldesormeau]

Hi All, resurrecting this thread to get opinions on having all L-type spectrometers compared to keeping a regular TAP/PET.

Potential JEOL config:
Sp1: TAPL, LDE1L (GPC)
Sp2: PETL, LDE6L (GPC)
Sp3: TAP, PET (GPC)
Sp4: PETL, LiFL (Xe)
Sp5: PETL, LiFL (Xe)

Concern with all L-type being extremely high count rate when analyzing silicates at high current. Any foreseeable issues with switching to lower intensity 2^nd & 3^rd order lines as suggested to us a way to justify all L-type? JEOL has been nudging us away from a 4xtal spectrometer, hence the regular TAP/PET alone and the L-type LDE1 and LDE6 for sensitivity on light element work. We will have a mix bag of geological and materials samples, focusing on primarily geological materials for trace elements and PGE analyses. Thanks for any insight

[sem-geologist]

Concern with all L-type being extremely high count rate when analyzing silicates at high current. Any foreseeable issues with switching to lower intensity 2^nd & 3^rd order lines as suggested to us a way to justify all L-type?

Exactly! Large crystals are not problem-free and have the drawbacks.

We here in Warsaw University have 2 microprobes SX100 and SXFiveFE.

SX100:
Sp1: LiF, PET, TAP, PC0 (low pressure)
Sp2: PET, LiF (high pressure)
Sp3: TAP, PC2 (low pressure)
Sp4: LLiF, LPET (high pressure; added as an upgrade, made analysis more pleasant on that machine)

SxFiveFE:
Sp1: LTAP, LPC0 (extended spectrometer, low pressure)
Sp2: LLiF, LPET (high pressure)
Sp3: LLiF, LPET (high pressure)
Sp4: TAP, PC3, PC1, PC2 (low pressure)
Sp5: LLiF, LPET (high pressure)

With SXFiveFE there is one huge drawback: switching between analytical conditions back and forth is troublesome. And thus doing high current analysis and having no small PET hurts, as LPET with major element analysis suffers a lot from pile-up type of "dead-time" (I would call it rather blind time). We are doing a lot of complicated geological and geo-experimental minerals (sparingly some material science analysis) often containing 30+ elements (once I had hit the limit of Cameca Peaksight with 40 elements, took me two days to find out why could not do analysis with 41 elements). LTAP is not practical at all for such materials as background measurement is simply impossible to do due to wider peaks compared to TAP (even normal TAP suffers from that, and on our machine TAP (Sp4) is blanked from sides to make peaks more narrow). LPC0 is nice for N,O,F, but I think PC0 would do not so much worse. I think I would be more happy with getting rid of double-turret with LTAP and LPC0 and get quadruple turret (which can't mount large crystals) with TAP, PET and PC0 (and probably move some other PC* from spectrometer 4).

As for SX100 - there is no big issue with the spectrometer configuration as the analytical conditions can be switched without drawbacks back and forth, and so there is no problem with running into pileup problem, simply by keeping the current respectively low.

BTW, pileup problem could be reduced with better pre-amplifiers and counting electronics.

[Mike Matthews]

What are JEOL users experience/opinions for L versus H crystals?

[Probeman]

What are JEOL users experience/opinions for L versus H crystals?

Hi Mike,
John Armstrong recommends against using TAPH crystals for quantitative work, which due to their extreme curvature (100 mm Rowland circle), often show cracks and other degradation, which results in spurious spectral artifacts.

Here is a topic on the subject:

https://probesoftware.com/smf/index.php?topic=611.0

[Probeman]

Hi All, resurrecting this thread to get opinions on having all L-type spectrometers compared to keeping a regular TAP/PET.

Potential JEOL config:
Sp1: TAPL, LDE1L (GPC)
Sp2: PETL, LDE6L (GPC)
Sp3: TAP, PET (GPC)
Sp4: PETL, LiFL (Xe)
Sp5: PETL, LiFL (Xe)

Concern with all L-type being extremely high count rate when analyzing silicates at high current. Any foreseeable issues with switching to lower intensity 2^nd & 3^rd order lines as suggested to us a way to justify all L-type? JEOL has been nudging us away from a 4xtal spectrometer, hence the regular TAP/PET alone and the L-type LDE1 and LDE6 for sensitivity on light element work. We will have a mix bag of geological and materials samples, focusing on primarily geological materials for trace elements and PGE analyses. Thanks for any insight

It's a legitimate concern if you are analyzing major elements using beam conditions for trace elements.  But what we normally do is run two conditions as seen here:



In this example (from a run by Karsten Goemann at the University of Tasmania), the major elements are acquired at 30 nA, and then the conditions are switched to 300 nA for the traces.  This is often done not only because of high count rates/deadtimes for major elements, but also to avoid beam damage effects for the major elements (regardless of the Bragg crystals utilized).

Of course, as mentioned by sem-geologist, this could be probematic for the SXFive instrument, which apparently takes time to stabilize after changing conditions, but Cameca Sx100 instruments and all JEOL instruments should have no issues with this type of acquisition.

I personally would definitely get as many large area crystals as I possibly could on a new instrument, the only caveat would be that one can't get large area crystals on 4 crystal spectrometers.

Finally, because sem-geologist mentioned the 40 element limit in PeakSight, I should mention that Probe for EPMA can handle up to 72 elements...