[Ifeffit] CN and bond distances in Artemis

[Anatoly Frenkel]

Dear JA,

"For the record", I don't recall saying anything to the effect that S02 =
0.2 "is close to the case of Pt nanoparticles" - which would be wrong since
the smallest number for 13 atom particles is that of the cuboctahedron:
5.54. Thus, the ratio of 5.54 and 12 makes it 0.46. That means, if S02 in
the foil is 0.85, and in the nanoparticle it is 0.2 (I assume your
degeneracy is 12 as in the foil), it corresponds to the coord. number of
(0.2/0.85)*12 = 2.8 which is too small to correspond to a sample with the
majority of well defined nanoparticles in it.

I can only assume (because it is not clear) that your "0.2" was obtained by
assuming 12 neighbors as in the foil, instead of doing what Matt suggested
to do (amp = S02*N, where N is variable).

Anatoly



-----Original Message-----
From: ifeffit-bounces at millenia.cars.aps.anl.gov
[mailto:ifeffit-bounces at millenia.cars.aps.anl.gov]On Behalf Of Juan
Antonio Macia Agullo
Sent: Tuesday, December 12, 2006 12:25 PM
To: XAFS Analysis using Ifeffit
Subject: Re: [Ifeffit] CN and bond distances in Artemis



Hi Matt,

Thank you very much for your comments. S02 is a variable for me and it gives
information about the size of the particles (related to coordination
number). For
instance, in the case of Pt foil it should be close to 0.85, however it is
much
smaller (0.20) in the case of platinum nanoparticles (size < 2nm), Anatoly
commented
that it is normal this low value in the case of nanoparticles. Probably,
what you
purposed it has more sense than what I did. Then, as I understand, I have to
suppose
reasonable value of S02, lets say 0.85, or shall I have to use S02 from Pt
foil?
Secondly, I define the variable amp = S02*N and, shall I write a value of
one in the
box corresponding to N? I mean, the sight in Artemis shows like this: N:1 X
S02:amp.

Talking about the Nyquist theorem, a higher number of independent points
implies a
higher number of variables that can be used and it is especially useful for
me when
I fit only the first shell (R-range is small) due to the low number of
independent
points obtained. If I use the formula "+2", I can define more variables. I
usually
use the crude rule: number of variables < 2/3 Nip. I read this "+2" formula
described by D.C. Koningsberger et al (Topics in Catalysis 10 (2000)
143-155) and I
though that was ok. What is the origin of this "+2"?

I have obtained a value of 0.53 for S02 in the case of Pt foil, it suggests
(as
people from this e-mailing list said) that a correction of self absorption
should be
used until reaching a reasonable value for S02 (0.85). After this, I will
apply the
same correction to the samples because they were measured in fluorescence
way too.
My samples contain less than 1% of platinum in weight (about 0.80%) and I do
not
know if there will be self-absorption.

Finally, I would like to ask about the detection of low Z scatterers
(usually O or C
from the support). I read in the aforementioned paper the difference file
technique,  Artemis includes this possibility? How can low Z scatterers be
detected?

Thanks a lot

Best regards,
JA