[Ifeffit] Error in Running XANES on FEFF8.4 for Cobalt based cluster

[Bruce Ravel]

On 07/29/2015 06:01 PM, Ganesh Subramanian wrote:
> 1. *Using the FOLP card*: The actual cluster size is about 180 atoms
> extending upto 18 A in radial distance from the central cobalt atom. But
> there are quite a few instances of Hydrogen atoms that are too close. I
> tried the FOLP with the recommendation for Hydrogen of 0.8, I seem to
> get a syntax type error, that i cant figure out. I checked the crystal
> structure and the smallest hydrogen bond distance is 0.89 A. (I have
> attached the corresponding input and log file with suffix 'full cluster')

In Feff the "TWO ATOMS VERY CLOSE TOGETHER" message gets shouted at you 
when you have atoms that are separated by 1.75 Angstroms or less.  That 
message has nothing to do with the FOLP setting.

> 2. *SCF for small cluster*: Since the problem with the hydrogen
> distances persisted, I decided to run the calculation for a smaller
> cluster (about 50 atoms). I used the conditions as prescribed for small
> clusters - SCF 5.0 1, FMS 5.0 1, RPATH -1. Those hydrogens that were too
> close (within this smaller cluster, there were about 3 of them, and none
> in the first or second shell of co-ordination) were removed manually.
> Now, I did not have the error on distances, did not require the FOLP
> card to be manually altered, but the log1 file seems to suggest that the
> SCF is not calculated. I cant figure why that is the case. (I have
> attached the corresponding input and log file with suffix 'small cluster')

When I ran feff 8.40, I got same the contents of the log1.dat file, but 
with these additional lines:

                 Core-valence separation
   WARNING: fatal error in subroutine corval. Try
     to reduce ca1 in SCF card. If does not help,
   SEND bug report to AUTHORS
   CORVAL-1

I don't quite know what that means, but it gives some hints.  The source 
code corval.f tells us that is "Finds the core-valence separation for 
the cluster of atoms."  I simply am not knowledgeable enough about the 
details of Feff to know why your cluster triggered this warning.

B



-- 
  Bruce Ravel  ------------------------------------ bravel at 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/