Suppose the mono is being used at the 333 reflection (for Si), intentionally working at the 3rd harmonic. In that case, if the fundamental isn't filtered out, you could get just what we see, with the sample doing the filtering before the beam gets to the transmission chamber. mam On 6/14/2023 9:16 PM, Anatoly Frenkel wrote:

Thank you, Matthew and Robert, but I think the explanation is more simplle. The fact that there are peaks in I0 and dips in transmission is a clue that explains why transmission is almost linear while I0 curves up.

Anatoly

...

On Jun 14, 2023, at 11:34 PM, matthew marcus wrote:

One thing that's pretty odd is that the glitches are peaks in I0, and dips in It. It would be much easier to explain if It was the one that increased steeply and I0 was more linear, but the presence of the edge in It shows that you didn't switch them. Red = It. What's that sudden drop at the end?

Could it be that the I0 chamber is misaligned so that beam is scraping one of the plates? That could cause all sorts of odd things to happen. mam

...

On 6/14/2023 5:27 PM, Anatoly Frenkel wrote: Hello, all. It is a low- to medium- level brain teaser. Pt-coated collimating mirror was in place for Pd K-edge measurement, but Au L3-edge of Pd-Au alloy was measured (for testing purposes). I0 and It detectors were both Ar filled ionization chambers. Because of the energy dependence of reflectivity of the Pt mirror, I0 intensity was strongly nonlinear (blue curve). However, the transmission intensity in the It detector was almost linear (red curve). Why? Anatoly image.png _______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit Unsubscribe: http://millenia.cars.aps.anl.gov/mailman/options/ifeffit

---

Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit Unsubscribe: http://millenia.cars.aps.anl.gov/mailman/options/ifeffit

_______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit Unsubscribe: http://millenia.cars.aps.anl.gov/mailman/options/ifeffit

Hi Anatoly, I think Robert or Matthew made this point, but if set up for Pd, the mirror angle may have been chosen to reject ~70 keV, but possibly not 36 keV -- the harmonic at the Au edge. Do you know what the mirror angle was? The Ar-filled I0 would be very efficient at absorbing 12 keV, and only pretty efficient at absorbing 36 keV. That would leave a more harmonic-rich beam exiting I0 and hitting the sample than entering I0. The good news is that the dense Pd/Au sample would be efficient at absorbing 36 keV too (but it was ~1 absorption length at 24 keV?) too. Mono reflectivity of 36 keV vs 12 keV might also factor in. If you were not deliberately detuning but the crystals were slightly misaligned, the harmonic content may change significantly over the scan range. I would not guess that to dominate, but maybe it factors in. On Thu, Jun 15, 2023 at 9:54 AM Anatoly Frenkel < anatoly.frenkel@stonybrook.edu> wrote:

Thank you, Matt. Ion chambers were filled with 90% Ar, and Pt coating was used, because we were measuring Pd K edge for the project, but we decided to look at the Au edge for testing purposes.

Anatoly

On Jun 15, 2023, at 10:45 AM, Matt Newville wrote:

 I'm not sure why the intensity would go up unless the ion chamber was poorly set up. But, as others have pointed out, the mirror reflectivity for a Pt mirror should not change significantly over this energy range - the energy range is not that close to the Pt L3 or L2 edges. Depending on where it was located, fluorescence from the Pt mirror might pollute the signal in the I0 ion chamber, but that would also likely be a fairly constant background.

But, why would you fill the I0 ion chamber with Argon? A 10-cm ion chamber filled with Ar would absorb about 50% of the beam at 12 keV. Even at 24 keV, that would absorb 8% of the beam - not necessarily a problem but also probably generating at least a micro-Amp, so way more signal than you would need.

For mirror reflectivity curves, allow me to humbly remind everyone of

https://xraydb.xrayabsorption.org/reflectivity/Pt/21.45/2.5/10/s/1000/51000/...

which is both interactive and works with X-ray energies above 30 keV.

On Wed, Jun 14, 2023 at 7:28 PM Anatoly Frenkel < anatoly.frenkel@stonybrook.edu> wrote:

...

Hello, all. It is a low- to medium- level brain teaser.

Pt-coated collimating mirror was in place for Pd K-edge measurement, but Au L3-edge of Pd-Au alloy was measured (for testing purposes). I0 and It detectors were both Ar filled ionization chambers. Because of the energy dependence of reflectivity of the Pt mirror, I0 intensity was strongly nonlinear (blue curve). However, the transmission intensity in the It detector was almost linear (red curve). Why?

Anatoly

_______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit Unsubscribe: http://millenia.cars.aps.anl.gov/mailman/options/ifeffit

-- --Matt Newville <newville at cars.uchicago.edu> 630-327-7411 _______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit Unsubscribe: http://millenia.cars.aps.anl.gov/mailman/options/ifeffit

_______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit Unsubscribe: http://millenia.cars.aps.anl.gov/mailman/options/ifeffit

-- --Matt Newville <newville at cars.uchicago.edu> 630-327-7411

HI Mike, it is dilute Pd in Au nanoparticles deposited on SiO2 with about 5% metal weight loading. In my view, the lack of curvature in transmission (the original problem in my post) is a simple consequence of two opposite trends, 1) the curving up of the transmission signal by the same reason (yet to be identified, which is something I will try to, after getting useful hints from many people here) as the curving up of the incident beam and 2) the opposite trend due to the greater removal of the photons at higher energy by I0 detector. The two opposite trends tend to cancel each other. Anatoly On Thu, Jun 15, 2023 at 4:14 PM Mike Massey wrote:

A question because I'm not really super clear on how harmonics work:

(I'll get to the question in a moment...) I note that I0 starts to go non-linear around 12200-12400 eV and stays that way. So an increase (proportional?) in the amount of signal in It might be happening at around the same point.

Could Pd fluorescence into It be impacting the signal in It? I realize that the Pd emission lines are very low in energy but they are around 3k eV.

I'm not super familiar with alloy systems but I did have an issue once upon a time with back fluorescence into an ion chamber and things went a bit wonky. This wouldn't exactly be "back" fluorescence, but if the system is mostly Pd with a little bit of Au, I guess...? And it might be easy enough to test, just put a high-pass filter in front of It?

I got nothin' but it's kinda fun to take a shot. Cheers,

Mike

On Jun 16, 2023, at 03:26, Matt Newville wrote:

 Hi Anatoly,

I think Robert or Matthew made this point, but if set up for Pd, the mirror angle may have been chosen to reject ~70 keV, but possibly not 36 keV -- the harmonic at the Au edge. Do you know what the mirror angle was?

The Ar-filled I0 would be very efficient at absorbing 12 keV, and only pretty efficient at absorbing 36 keV. That would leave a more harmonic-rich beam exiting I0 and hitting the sample than entering I0. The good news is that the dense Pd/Au sample would be efficient at absorbing 36 keV too (but it was ~1 absorption length at 24 keV?) too.

Mono reflectivity of 36 keV vs 12 keV might also factor in. If you were not deliberately detuning but the crystals were slightly misaligned, the harmonic content may change significantly over the scan range. I would not guess that to dominate, but maybe it factors in.

On Thu, Jun 15, 2023 at 9:54 AM Anatoly Frenkel < anatoly.frenkel@stonybrook.edu> wrote:

...

Thank you, Matt. Ion chambers were filled with 90% Ar, and Pt coating was used, because we were measuring Pd K edge for the project, but we decided to look at the Au edge for testing purposes.

Anatoly

On Jun 15, 2023, at 10:45 AM, Matt Newville wrote:

 I'm not sure why the intensity would go up unless the ion chamber was poorly set up. But, as others have pointed out, the mirror reflectivity for a Pt mirror should not change significantly over this energy range - the energy range is not that close to the Pt L3 or L2 edges. Depending on where it was located, fluorescence from the Pt mirror might pollute the signal in the I0 ion chamber, but that would also likely be a fairly constant background.

But, why would you fill the I0 ion chamber with Argon? A 10-cm ion chamber filled with Ar would absorb about 50% of the beam at 12 keV. Even at 24 keV, that would absorb 8% of the beam - not necessarily a problem but also probably generating at least a micro-Amp, so way more signal than you would need.

For mirror reflectivity curves, allow me to humbly remind everyone of

https://xraydb.xrayabsorption.org/reflectivity/Pt/21.45/2.5/10/s/1000/51000/...

which is both interactive and works with X-ray energies above 30 keV.

On Wed, Jun 14, 2023 at 7:28 PM Anatoly Frenkel < anatoly.frenkel@stonybrook.edu> wrote:

...

Hello, all. It is a low- to medium- level brain teaser.

Pt-coated collimating mirror was in place for Pd K-edge measurement, but Au L3-edge of Pd-Au alloy was measured (for testing purposes). I0 and It detectors were both Ar filled ionization chambers. Because of the energy dependence of reflectivity of the Pt mirror, I0 intensity was strongly nonlinear (blue curve). However, the transmission intensity in the It detector was almost linear (red curve). Why?

Anatoly

_______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit Unsubscribe: http://millenia.cars.aps.anl.gov/mailman/options/ifeffit

-- --Matt Newville <newville at cars.uchicago.edu> 630-327-7411 _______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit Unsubscribe: http://millenia.cars.aps.anl.gov/mailman/options/ifeffit

_______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit Unsubscribe: http://millenia.cars.aps.anl.gov/mailman/options/ifeffit

-- --Matt Newville <newville at cars.uchicago.edu> 630-327-7411 _______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit Unsubscribe: http://millenia.cars.aps.anl.gov/mailman/options/ifeffit

_______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit Unsubscribe: http://millenia.cars.aps.anl.gov/mailman/options/ifeffit