[Wien] Force optimization of a spinel compound

Laurence Marks L-marks at northwestern.edu
Mon Sep 8 21:12:44 CEST 2014


What you have done is probably "right", although a 1x1x1 supercell is small
and to do a formally correct calculation you would need to use a NxNxN
supercell with N large and randomly occupy the positions but this would
become very computationally expensive.

In a structure with partial occupancy, locally the atomic positions of your
spinel will deviate from (x,x,x), similar to what you are finding. What XRD
is measuring is a statistical average over the real local (x,y,z) which
leads to an average (x,x,x). I expect that if you look carefully you will
find that Shelx (or similar) is refining relatively large temperature
factors for the oxygens which are being displaced. If you were to do the
XRD at different temperatures you might find that the temperature factors
are not going to small numbers as T-->0, rather to some large number due to
the athermal disorder.

N.B., of course, you could also have a superstructure develop at lower
temperatures and/or with more annealing of your sample.

On Mon, Sep 8, 2014 at 12:58 PM, shamik chakrabarti <shamikphy at gmail.com>
wrote:

>  Dear Laurence Marks Sir,
>
>                            Yes I have converted the 14 atomic Fd3m cell
> in 56 in equivalent atoms cell (generated from 1x1x1 supercell with target
> lattice primitive) to incorporate partial occupancy. However, before
> starting optimization I have checked that implemented structure (used as
> input) provides the same XRD as the experimental structure with same
> relative intensities corresponding to different peaks. However that
> implemented structure shows a force as large as 200 mRy/bohr for some atoms
> and hence can not be used for electronic structure calculation & we need
> optimization. Yes, it is true that as I have not kept the space group
> constraint to incorporate the partial occupancies the oxygen coordinate
> transformed from (x,x,x) to (x,y,z) during optimization,
>
>  But, then what should we do in this scenario?.....
>
>  Looking forward to your advice.
>
>  Thanks in advance.
>
>  with regards,
>
> On Mon, Sep 8, 2014 at 11:15 PM, Laurence Marks <L-marks at northwestern.edu>
> wrote:
>
>> This is not a simple question to answer. In both DFT and XRD the symmetry
>> of the structure is an input parameter, and in both cases it is sometimes
>> wrong. The most obvious question to ask is whether you used the same Space
>> Group for your refinement in Wien2k that was used in the XRD refinement. I
>> have a suspicion that you did not, as the presence of (x,x,x) positions
>> implies a Space Group constraint and if you are getting (x,y,z) then you
>> probably have a different symmetry in Wien2k.
>>
>> Check first that the symmetry is completely the same.
>>
>>  If they are the same, the question then would be why one gets (x,x,x)
>> in the XRD refinement? It could be disorder or something else, but without
>> more information it is hard to guess.
>>
>>
>> On Mon, Sep 8, 2014 at 11:37 AM, shamik chakrabarti <shamikphy at gmail.com>
>> wrote:
>>
>>>  Dear wien2k users,
>>>
>>>            We have optimized atomic coordinates of a spinel cubic
>>> compound containing 56 atoms/cell. The most important variation we have
>>> seen between experimental structural coordinates & DFT (GGA+U) optimized
>>> coordinates is that.
>>>
>>>  the oxygen coordinate changes from (x,x,x) configuration to (x,y,z)
>>> configuration as we have not constrained the oxygen coordinates during
>>> optimization.
>>>
>>>  Although, this change has generated a structure having very small
>>> forces (< 1 mRy/Bohr) per atoms, the change in oxygen coordinates are
>>> significant (difference between x & z is 0.021) enough to change the
>>> relative intensities of different peaks of XRD patterns.
>>>
>>>  Now my query is (before starting the experimental simulation by myself
>>> again!!) whether there is a way to constrain atomic coordinates such that
>>> it remain constrained to (x,x,x) configuration while changing the value of
>>> x?
>>>
>>>  & also whether such a constrained way should be adopted or should we
>>> accept the DFT optimized structure (a bit different from experimental
>>> one!!) as it is for further calculation?
>>>
>>>  Looking forward to your opinions & suggestions.
>>>
>>>  Thanks in advance,
>>>
>>>  with regards,
>>>
>>> --
>>> Shamik Chakrabarti
>>> Senior Research Fellow
>>> Dept. of Physics & Meteorology
>>> Material Processing & Solid State Ionics Lab
>>> IIT Kharagpur
>>> Kharagpur 721302
>>> INDIA
>>>
>>
>>
>>
>>   --
>> Professor Laurence Marks
>> Department of Materials Science and Engineering
>> Northwestern University
>> www.numis.northwestern.edu
>> Corrosion in 4D: MURI4D.numis.northwestern.edu
>> Co-Editor, Acta Cryst A
>> "Research is to see what everybody else has seen, and to think what
>> nobody else has thought"
>> Albert Szent-Gyorgi
>>
>> _______________________________________________
>> Wien mailing list
>> Wien at zeus.theochem.tuwien.ac.at
>> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>> SEARCH the MAILING-LIST at:
>> http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html
>>
>>
>
>
>  --
> Shamik Chakrabarti
> Senior Research Fellow
> Dept. of Physics & Meteorology
> Material Processing & Solid State Ionics Lab
> IIT Kharagpur
> Kharagpur 721302
> INDIA
>



-- 
Professor Laurence Marks
Department of Materials Science and Engineering
Northwestern University
www.numis.northwestern.edu
Corrosion in 4D: MURI4D.numis.northwestern.edu
Co-Editor, Acta Cryst A
"Research is to see what everybody else has seen, and to think what nobody
else has thought"
Albert Szent-Gyorgi
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20140908/1605fc83/attachment.htm>


More information about the Wien mailing list