U Ma verses Mb lines

[wrigke]

In doing a wavescan on uranium, I noticed that the intensity of Uma is considerably less than that of Umb.  This isn't just me--there are published scans that show the same thing (e.g. Walker, 1999, taken from Kleykamp)

There is an Ar absorption edge to the high sine side of Uma, but can that account for such a large difference in peak intensity (i.e. 90cps for Umb versus 50cps for Uma)? 

Cheers,

Karen

[Mike Jercinovic]

Hi Karen,
Yes, indeed.  The Ar edge is actually on the short wavelength side of U-Ma, so between UMa and UMb.  The height of this edge depends on the pressure of the gas in the counter tube, but is clearly visible below 3 bars.  Generally the high P tubes on the SX are about 2 bars, and low P counters are around 1 bar, where the edge is most strongly expressed.  The amount of Th obviously affects the relative peak heights also as the ThMb-UMa interference is very large compared to ThMg-UMb.  The net result is that the count rates for U Mb are generally similar to UMa, but UMb is better resolved (still have to use a good interference correction though!).  So the actinide M series x-ray spectrum is an interesting study in emission probabilities (mostly N to M transitions) and counter gas quantum efficiencies.  Depending on the beam current and kV, U and Th concentrations, and Ar edge height, a difference of 90cps for UMb compared to 50cps on UMa is certainly possible.

Mike J.

[John Donovan]

Mike is correct.

The other reason the U Mb line can be a higher intensity is that the U Mb line can ionize the Ar K edge, but the Ma line cannot!  So the better count rate for U Mb is because it is absorbed more efficiently by the detector gas depending on the pressure as Mike points out.

See the attached graph here which shows the U Ma line to be higher, but only because it is sitting on a large Th peak shoulder.  Remember that one needs to be logged in to see attachments!

[Julien]

Hi Karen,

Mike J is always correct Following our discussion last week, I did perform some scan over the U Ma-Mb region on my U metal standard using PET on Ar counter versus Xe counter (just to get a HUGE amount of counts ) and I got this results. I think this PDF speak for itself. The (rough) peak ratio is indicated on the graphic, and effectively the M-alpha peak is much smaller than M-beta for the reason Mike explained. However this effect is definitely not present on the Xe-counter, further confirming the dramatic effect of the Ar-absorption edge.

Note that I performed these analyses at 25 kV, but I expect the results at 15 kV to be similar. Maybe we should suggest Cameca to build some Xe counter

Best,

Julien

[Mike Jercinovic]

Just checking back in on this one.  Note also that this is why Cameca uses variable counter tube pressure so as to increase the quantum efficiency of P10 for more energetic lines.  So, the 1bar P10 performance produces UMa< UMb, but increasing the pressure to 2 bars give UMa>UMb, similar to the sealed Xe counter (see attached).  Note this plot is from UO2 at 15kV.  The Ar edge height is also reduced at higher counter tube pressure.

Mike J.

[Probeman]

And for those that haven't dealt with this before, here's the Ar absorption edge in YPO4 in the U M a/b region.



Ultimately we need to use an LIF220 and measure the U La line (1st order) using a tandem gas and solid state detector like this



for best moderate to high energy sensitivity. Someday...

[Probeman]

So here is a recent example (uranium k-alpha in ZrO2) of me being quite stupid and using the multi-point background defaults without checking carefully :



Fortunately we can modify the background fitting in post processing, and force the multi-point background to only use the background measurements on one side of the absorption edge to get a more accurate measurement of the intensity under the peak next to the Ar absorption edge:



The small peak shown here is approx. 2200 PPM:

[Philipp Poeml]

Hi there,

I have a question about this. This is how a spectrum of UO2 looks like on our 1011 Qtz crystal high pressure spectro (2 bars P10).




So we have the Ar absorption edge on the short wavelength side of the U Ma peak.

1) how does this work with this stupid edge? I mean, in the sense of: Looking at the other posts in this thread, it affects also the background intensity. Is that the case for the whole "right" side of the spectrum? Will the intensity of the background be lower than on the "left" side for the rest of the spectrum? I am asking this, because I am thinking about where to measure the background for U Ma here. It is clearly super high on the short wavelength side.

2) So wouldn't it be a good idea to use the multi point background method? Like shown in the image? I would measure the backgrounds left and right of the whole are of U M lines and put some exponential fit through. However, what about this absorption edge then? Imagine how this area looks like in a sample where we have U, Np, Pu, and Am. It's like the REEs, just worse. I think the multipoint option is the only reasonable approach, whatever there is with the absorption edge of Ar.

3) Another question in this respect: Is it possible to correct for the Ar absorption? Let's say I would know exactly the ratio of the U Ma-Mb second order, would that be the same ratio as for the U Ma-Mb first order? Or would it be better to compare to the ratios of Np or Pu Ma-Mb? Or, in addition, could I "correct" for the Ar edge to bring the Th Ma line up to a level where I could compare it to Np Ma or Pu Ma?

What do you think?

Cheers
Ph