Effect of gradual thickness variation in beam
Sample cell Hi all, I'm planning on some transmission-mode XAS with smaller than usual sample tubes. I'm realizing I might be in danger of creating excessive non-uniformity in my samples by having the beam size on the same order of magnitude as the sample tube radius. For example, let's say I want to measure liquid in a sample tube with a 1000 micron outer diameter, with a beam 500 microns wide and centered on the sample tube. If the tube is orthogonal to the ring plane, the entire vertical portion of the beam will pass through the same length of liquid. But in the horizontal plane, the center of the beam will pass through 1000 microns of sample while the edges of the beam will pass through 866 microns of sample, due to the curvature of the sample cell across the horizontal plane. Most of what I know about the statistics of thickness effects are about leakage and pinholes - nonlinearity caused by a few spots having very low or negligible sample thickness. But I don't know how significant a "mild" thickness distribution might be. If this sort of thickness distribution is going to be an issue it would be great to know that beforehand and either go with larger samples, smaller beam size, or more creative orientation. Thanks, Jason
Hi Jason, You are saying that the outer diameter is 1000 um and that the center of the beam will pass through 1000 micron of the sample... How possible? The tube is not infinitely thin... The sensitivity of transmission XAS to the non-uniformity of the sample will be different at different energies. It is important to know how absorbing is the sample (total mu) and what is the delta mu for the X-ray energy range you are interested in. Anatoly ________________________________ From: Ifeffit [ifeffit-bounces@millenia.cars.aps.anl.gov] on behalf of Jason Gaudet [jason.r.gaudet@gmail.com] Sent: Wednesday, January 27, 2016 7:05 PM To: XAFS Analysis using Ifeffit Subject: [Ifeffit] Effect of gradual thickness variation in beam Sample cell Hi all, I'm planning on some transmission-mode XAS with smaller than usual sample tubes. I'm realizing I might be in danger of creating excessive non-uniformity in my samples by having the beam size on the same order of magnitude as the sample tube radius. For example, let's say I want to measure liquid in a sample tube with a 1000 micron outer diameter, with a beam 500 microns wide and centered on the sample tube. If the tube is orthogonal to the ring plane, the entire vertical portion of the beam will pass through the same length of liquid. But in the horizontal plane, the center of the beam will pass through 1000 microns of sample while the edges of the beam will pass through 866 microns of sample, due to the curvature of the sample cell across the horizontal plane. Most of what I know about the statistics of thickness effects are about leakage and pinholes - nonlinearity caused by a few spots having very low or negligible sample thickness. But I don't know how significant a "mild" thickness distribution might be. If this sort of thickness distribution is going to be an issue it would be great to know that beforehand and either go with larger samples, smaller beam size, or more creative orientation. Thanks, Jason
On 01/27/2016 07:05 PM, Jason Gaudet wrote:
I'm planning on some transmission-mode XAS with smaller than usual sample tubes. I'm realizing I might be in danger of creating excessive non-uniformity in my samples by having the beam size on the same order of magnitude as the sample tube radius.
For example, let's say I want to measure liquid in a sample tube with a 1000 micron outer diameter, with a beam 500 microns wide and centered on the sample tube. If the tube is orthogonal to the ring plane, the entire vertical portion of the beam will pass through the same length of liquid. But in the horizontal plane, the center of the beam will pass through 1000 microns of sample while the edges of the beam will pass through 866 microns of sample, due to the curvature of the sample cell across the horizontal plane.
Most of what I know about the statistics of thickness effects are about leakage and pinholes - nonlinearity caused by a few spots having very low or negligible sample thickness. But I don't know how significant a "mild" thickness distribution might be. If this sort of thickness distribution is going to be an issue it would be great to know that beforehand and either go with larger samples, smaller beam size, or more creative orientation.
Jason, There may be some distortion to the data due to the varying thickness. On the plus side, liquids tend to be very homogeneous. Probably the best solution to this situation, if it's available at the beamline, is to focus the beam in the vertical, for example, using a flat mirror on a bender. Focusing to a spot would also address the thickness situation, but you need to be mindful that the intense beam could generate radicals from the liquid, causing sample damage. If the sample is sufficiently concentrated, then you could simply slit the beam down to 100 or 200 microns and still have enough for a transmission measurement. In fact, that would be the way to decide if you are seeing a big problem from the shape of the tube. Compare the mu(E) with the slits at 500 to mu(E) with the slits at 100 (carefully aligning the sample both times, of course). If they are the same, then, then you're golden. B -- Bruce Ravel ------------------------------------ bravel@bnl.gov National Institute of Standards and Technology Synchrotron Science Group at NSLS-II Building 535A Upton NY, 11973 Homepage: http://bruceravel.github.io/home/ Software: https://github.com/bruceravel Demeter: http://bruceravel.github.io/demeter/
Hi Anatoly,
Sorry, I meant to write 1000 um inner diameter.
Energies will be V and Fe, 5465 & 7112 eV. Samples are roughly 1
absorption length thick for the metal (V, Fe) and about 2-2.5 absorption
lengths for the entire sample.
Bruce,
I may just flip over the sample holder and have the sample tubes in plane
with the storage ring to take advantage of the greater horizontal spread,
rather than focusing on the vertical. Cutting the vertical slits down to
100-300 um while keeping the horizontal around 500 um ought to get me the
most beam at the least variation in mu(E), especially if I go with 1500 um
ID tubes. The samples are nice and concentrated so I can afford to give up
a lot of photon flux. Bar-napkin calculations tell me with a 1500 um ID
tube and 200 um V 750 um H slits I'll lose an acceptable amount of photon
flux and have a <1% thickness variation. I'm fairly confident I've made
pressed pellets with as much or more variation from one spot to another,
without noticeable mu(E) changes from spot to spot, in the vicinity of this
energy range.
I'll go ahead and take your advice, and amend my setup procedure with some
measurements of Io intensity and mu(E) reproducibility as a function of
slit width. While a bit of a bother, it's still worth it to keep our
sample prep simple and inexpensive.
Thanks,
Jason
On Thu, Jan 28, 2016 at 9:12 AM, Bruce Ravel
On 01/27/2016 07:05 PM, Jason Gaudet wrote:
I'm planning on some transmission-mode XAS with smaller than usual sample tubes. I'm realizing I might be in danger of creating excessive non-uniformity in my samples by having the beam size on the same order of magnitude as the sample tube radius.
For example, let's say I want to measure liquid in a sample tube with a 1000 micron outer diameter, with a beam 500 microns wide and centered on the sample tube. If the tube is orthogonal to the ring plane, the entire vertical portion of the beam will pass through the same length of liquid. But in the horizontal plane, the center of the beam will pass through 1000 microns of sample while the edges of the beam will pass through 866 microns of sample, due to the curvature of the sample cell across the horizontal plane.
Most of what I know about the statistics of thickness effects are about leakage and pinholes - nonlinearity caused by a few spots having very low or negligible sample thickness. But I don't know how significant a "mild" thickness distribution might be. If this sort of thickness distribution is going to be an issue it would be great to know that beforehand and either go with larger samples, smaller beam size, or more creative orientation.
Jason,
There may be some distortion to the data due to the varying thickness. On the plus side, liquids tend to be very homogeneous.
Probably the best solution to this situation, if it's available at the beamline, is to focus the beam in the vertical, for example, using a flat mirror on a bender. Focusing to a spot would also address the thickness situation, but you need to be mindful that the intense beam could generate radicals from the liquid, causing sample damage.
If the sample is sufficiently concentrated, then you could simply slit the beam down to 100 or 200 microns and still have enough for a transmission measurement. In fact, that would be the way to decide if you are seeing a big problem from the shape of the tube. Compare the mu(E) with the slits at 500 to mu(E) with the slits at 100 (carefully aligning the sample both times, of course). If they are the same, then, then you're golden.
B
-- Bruce Ravel ------------------------------------ bravel@bnl.gov
National Institute of Standards and Technology Synchrotron Science Group at NSLS-II Building 535A Upton NY, 11973
Homepage: http://bruceravel.github.io/home/ Software: https://github.com/bruceravel Demeter: http://bruceravel.github.io/demeter/ _______________________________________________ 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
participants (3)
-
Anatoly I Frenkel
-
Bruce Ravel
-
Jason Gaudet