Dear all,
I would like to submit to your attention a problem with Atoms that I
still do not understand. What I would like to do is to generate a FEFF
input file starting from a supercell, that is repeating the unit cell
NXxNYxNZ times and create the crystal with the resulting box.
An example is given in the attached _test-atoms.inp_ created following
this procedure: from wurtzite GaN (a,c cell parameters) that we describe
with 2 atoms I reduce the symmetry and I use 4 atoms, then I create a 72
atoms supercell - 4*(3x3x2) - that has 3ax3ax2c dimensions and finally I
tell atoms to use "P1" space group in order to replicate the supercell
in x,y,z space and shift at the center the absorbing atom.
Here the *problem*: the resulting _test-feff8.inp_ present good values
of distances but strangely it has generated an additional Fe potential
(there is just 1 Fe absorbing atom in the box/list). This is annoying
because I have to remove manually the wrong "ipot" to run FEFF.
Is it a stupid mistake generating the Atoms input file or is it an odd
behaviour?
Cheers,
Mauro
PS: Atoms 3.0.1 (linux) 05/08/15
PPS: Do not pay attention to coordinates and distances, these are just
test-files!
--
Mauro Rovezzi
Hi Mauro,
If I understand correctly, you would like to generate a feff.inp file from an atoms.inp file where the only Fe atom is the absorber. But the resulting feff.inp file can not be run because it always makes an ipot 1 for a neighboring Fe atom but there are not any. So you have to go in by hand and remove that ipot number and shift the list because the ipot numbers have to go in order.
Yes, this is a feature and not so much a bug in the way that atoms works. Atoms assumes there will always be at least on Fe neighbor. It cuts some time by moving the last potential in the ipot list to number 1 so that you don't have to change all of them. You can also cheat by adding an Fe neighbor to the end that is not used and not close to any of the paths that you do use.
Cheers,
Shelly
________________________________
From: ifeffit-bounces@millenia.cars.aps.anl.gov on behalf of Mauro Rovezzi
Sent: Fri 9/19/2008 7:40 AM
To: Ifeffit-ML
Subject: [Ifeffit] [ATOMS] supercell input file
Dear all,
I would like to submit to your attention a problem with Atoms that I
still do not understand. What I would like to do is to generate a FEFF
input file starting from a supercell, that is repeating the unit cell
NXxNYxNZ times and create the crystal with the resulting box.
An example is given in the attached _test-atoms.inp_ created following
this procedure: from wurtzite GaN (a,c cell parameters) that we describe
with 2 atoms I reduce the symmetry and I use 4 atoms, then I create a 72
atoms supercell - 4*(3x3x2) - that has 3ax3ax2c dimensions and finally I
tell atoms to use "P1" space group in order to replicate the supercell
in x,y,z space and shift at the center the absorbing atom.
Here the *problem*: the resulting _test-feff8.inp_ present good values
of distances but strangely it has generated an additional Fe potential
(there is just 1 Fe absorbing atom in the box/list). This is annoying
because I have to remove manually the wrong "ipot" to run FEFF.
Is it a stupid mistake generating the Atoms input file or is it an odd
behaviour?
Cheers,
Mauro
PS: Atoms 3.0.1 (linux) 05/08/15
PPS: Do not pay attention to coordinates and distances, these are just
test-files!
--
Mauro Rovezzi
On Friday 19 September 2008 11:07:22 Kelly, Shelly wrote:
You can also cheat by adding an Fe neighbor to the end that is not used and not close to any of the paths that you do use.
That may be easier than my work-around. Good one! B -- Bruce Ravel ------------------------------------ bravel@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/
Bruce Ravel wrote:
On Friday 19 September 2008 11:07:22 Kelly, Shelly wrote:
You can also cheat by adding an Fe neighbor to the end that is not used and not close to any of the paths that you do use.
That may be easier than my work-around. Good one!
Yes, $ atoms -8 -r 10.5 test-atoms.inp but I get this FEFF error: Feff 8.40 XANES: name: Fe in GaN TI notes1: 3x3x2 GaN supercell Calculating potentials ... free atom potential and density for atom type 0 free atom potential and density for atom type 1 free atom potential and density for atom type 2 free atom potential and density for atom type 3 initial state energy overlapped potential and density for unique potential 0 overlapped potential and density for unique potential 1 overlapped potential and density for unique potential 2 overlapped potential and density for unique potential 3 muffin tin radii and interstitial parameters FOLP for POTENTIAL type 0 is too big. Reduce overlap using FOLP and rerun MOVRLP-1 It seems a FEFF story... I need to give a look at the CARDS... ;) Anyway, thanks for your previous answers! M
Mauro, That might have to do with N atom at 1.05 A. What happens if you simply remove that atom from the feff.inp file? B On Friday 19 September 2008 11:26:57 Mauro Rovezzi wrote:
Bruce Ravel wrote:
On Friday 19 September 2008 11:07:22 Kelly, Shelly wrote:
You can also cheat by adding an Fe neighbor to the end that is not used and not close to any of the paths that you do use.
That may be easier than my work-around. Good one!
Yes,
$ atoms -8 -r 10.5 test-atoms.inp
but I get this FEFF error:
Feff 8.40 XANES: name: Fe in GaN TI notes1: 3x3x2 GaN supercell Calculating potentials ... free atom potential and density for atom type 0 free atom potential and density for atom type 1 free atom potential and density for atom type 2 free atom potential and density for atom type 3 initial state energy overlapped potential and density for unique potential 0 overlapped potential and density for unique potential 1 overlapped potential and density for unique potential 2 overlapped potential and density for unique potential 3 muffin tin radii and interstitial parameters FOLP for POTENTIAL type 0 is too big. Reduce overlap using FOLP and rerun MOVRLP-1
It seems a FEFF story... I need to give a look at the CARDS... ;)
Anyway, thanks for your previous answers!
M _______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit
-- Bruce Ravel ------------------------------------ bravel@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/
Mauro, Your problem is a genuine bug. When I go through the contents of the cell to determine how many ipots to assign, the Fe atom gets noticed and added to the list of species in the cell. I should exclude the absorber from consideration at that step. I don't see immediately how to fix that problem, but I'll try to come up with a solution and check it into svn soon. I can suggest a work-around. It still requires editing the feff.inp file generated, but it is a much less onerous editing chore: * In the atoms.inp file, change the iron atom to Ga and give it a tag -- something like this: Ga 0.444444444444444 0.555555555555555 0.3455528537 Abs * Change the value of the "core" keyword to the tag: core = Abs * Run feff, and edit the first element of the potentials list. That is still annoying, but much less so than the feff.inp file you attached to your email. B On Friday 19 September 2008 10:40:04 Mauro Rovezzi wrote:
Dear all,
I would like to submit to your attention a problem with Atoms that I still do not understand. What I would like to do is to generate a FEFF input file starting from a supercell, that is repeating the unit cell NXxNYxNZ times and create the crystal with the resulting box.
An example is given in the attached _test-atoms.inp_ created following this procedure: from wurtzite GaN (a,c cell parameters) that we describe with 2 atoms I reduce the symmetry and I use 4 atoms, then I create a 72 atoms supercell - 4*(3x3x2) - that has 3ax3ax2c dimensions and finally I tell atoms to use "P1" space group in order to replicate the supercell in x,y,z space and shift at the center the absorbing atom.
Here the *problem*: the resulting _test-feff8.inp_ present good values of distances but strangely it has generated an additional Fe potential (there is just 1 Fe absorbing atom in the box/list). This is annoying because I have to remove manually the wrong "ipot" to run FEFF.
Is it a stupid mistake generating the Atoms input file or is it an odd behaviour?
Cheers, Mauro
PS: Atoms 3.0.1 (linux) 05/08/15 PPS: Do not pay attention to coordinates and distances, these are just test-files!
-- Bruce Ravel ------------------------------------ bravel@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/
participants (3)
-
Bruce Ravel -
Kelly, Shelly -
Mauro Rovezzi