[Ifeffit] "Hard tests failed in fovrg"

[Bruce Ravel]

On Friday 30 January 2009 08:49:36 am tony vitova wrote:
> Hi all,
>  
> I am trying to fit Eu L3 edge EXAFS data of Eu complex whose structure is
> known (hopefully correct). The default feff6l in Artemis reports an error
>  “Hard tests failed in fovrg. Muffin-tin radius may be too large;
> coordination number too small“. When I run feff8.2, this error exchanges
> with the following message “The phase shift correction is accurate to
> k=19”. I guess that this is not a problem, because my data is until k equal
> 12 A^-1. However, the fit of the first shell (9O/9N) gives Eo equal to 13,
> which is unrealistic. This fit with paths calculated with feff6l gives
> reasonable Eo value. But the phase shifts seem to be calculated
> incorrectly. I will be very thankful for any hint! Thanks in advance

Tonya,

A quit bit of googling turned up this useful mailing list post from
John Rehr from a couple of years ago about the "Hard test fail" thing
in feff6:

http://millenia.cars.aps.anl.gov/pipermail/ifeffit/2005-October/002080.html

That doesn't answer your question about feff8 however.  John or one of
his students would be better able to address that, however one might
be concerned that Feff has made a bad choice about which electrons to
consider as core electrons and which to consider as valence
electrons.  A bad choice there could confuse teh self-consistency
calculation.

I think the bottom line is whether the rest of the results besides E0
make physical sense....

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/