[Ifeffit] SCF and FMS radius

Feng Wang fwang at phys.ualberta.ca
Mon Jan 12 10:33:44 CST 2009

Bruce and Eugenio,

It is generally believed that the cluster size (or called spatial 
resolution) is mainly limited by MFP, to around 1 nm (in the near-edge 
region), but further spatial limitation may come from elastic scattering 
(i.e.  through phase cancellation). This is what we tried to understand 
through FEFF calculations in one paper. I put the abstract here in case you 
want to check: 

We have investigated the factors that determine the degree of localization 
of the information obtainable from electron energy loss or x-ray absorption 
fine structure. Inelastic scattering of the excited core electron limits the 
volume of specimen contributing to the backscattered intensity to a diameter 
in the range of 1–2 nm, dependent on the excited-electron energy and the 
composition and crystal structure of the sample. Phase cancellation between 
the backscattered waves further reduces the effective diameter that 
determines the observed fine structure to below 1 nm. Since the spatial 
resolution attainable by transmission electron microscopy can approach 0.2 
nm or can even be below 0.1 nm (with aberration correction), we predict that 
delocalization arising from the excited-electron range may limit the 
resolution of images based on changes in core-loss fine structure. (J. Appl. 
Phys. 104, 034906 (2008)).


> On Sunday 11 January 2009 12:23:26 pm Eugenio Otal wrote:
> > I am simulating the XANES spectra of a erbium atom in an interstitial 
> > of ZnO. I needed a SCF radius of 9.75 (316 atoms) and FMS radius of 11.90
> > (599 atoms) to obtain no change in the simulated spectra (differences of
> > 0.01% or less)
> > Now I am trying to simulate Er(OH)3 and it is using SCF 10.90 (622 atoms)
> > and I still haven found convergence, I will need more even larger radius
> > for FMS.
> > Does the XAS phenomena have this distances? Has this distances physical
> > meaning?
> Eugenio,
> The mean free path is quite long just above the Fermi energy.  It is 
> conceivable that, in a well ordered material, the cluster size required 
> convergence can be quite large.  I believe that a very large cluster -- 
> something like 13 A -- is needed for pure silicon.  Of course, I doubt 
> erbium hydroxide is that well-ordered!
> So, are you are seeing physics associated with the mean free path or a 
> in Feff's construction of the potentials?  I'm afraid I don't know how to 
> answer that, but hopefully John or one of his gang will weigh in.  Do you 
> the changes larger than your threshold over the entire data range or only 
> very close to the edge?
> Good luck,
> B
> -- 
>  Bruce Ravel  ------------------------------------ bravel at bnl.gov
>  National Institute of Standards and Technology
>  Synchrotron Methods Group at NSLS --- Beamlines U7A, X24A, X23A2
>  Building 535A
>  Upton NY, 11973
>  My homepage:    http://xafs.org/BruceRavel
>  EXAFS software: http://cars9.uchicago.edu/~ravel/software/exafs/
> _______________________________________________
> Ifeffit mailing list
> Ifeffit at millenia.cars.aps.anl.gov
> http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit

National Institute for Nanotechnology
11421 Saskatachewan Drive
Edmonton, Canada T6G 2M9

More information about the Ifeffit mailing list