calculation of a x-ray structure

[Jörg Saßmannshausen]

Dear all,

when I am usually calculate structures from the data I am obtaining via a x-
ray analysis, i.e. from the CIF file, I simply take the single molecule and run
that via some DFT functional. No periodic boundries, if there are two
molecules in the unit cell I simply pick one.

However, to find out whether a specific effect I am seeing is due to crystal
packing or not, I actually want to calculate the complete unit cell, which
contains two molecules, and I want to calculated it as a lattice, i.e. one
more cell in each direction. This way, I hope, I would see whether it is due
to crystal packing effects what I am observing, or not.

From the CIF file I know it is a triclinic cell and I also know the geometric
parameters, i.e. a,b,c and the angles.

If I plug that into the subsys section of the input file and I am using the MT
POISSON_SOLVER, what else do I need to bear in mind?
Does that actually make sense what I am trying to do or is that a waste of my
time? I never done solid state calculations as I am more a single molecule
person so your advice here would be much appreciated.

Please let me know if you got any further questions.

All the best from a dull London

Jörg

--
*************************************************************
Dr. Jörg Saßmannshausen, MRSC
University College London
Department of Chemistry
20 Gordon Street
London
WC1H 0AJ

email: j.sassma...@ucl.ac.uk
web: http://sassy.formativ.net

Please avoid sending me Word or PowerPoint attachments.
See http://www.gnu.org/philosophy/no-word-attachments.html

[hut...@chem.uzh.ch]

Hi

I would just do a periodic calculation (using the cif file).
You get all the packing effects and the computational costs
are not higher (or even lower) than for a cluster calculation.

regards

Juerg
--------------------------------------------------------------
Juerg Hutter                         Phone : ++41 44 635 4491
Institut für Chemie C                FAX   : ++41 44 635 6838
Universität Zürich                   E-mail: hut...@chem.uzh.ch
Winterthurerstrasse 190
CH-8057 Zürich, Switzerland
---------------------------------------------------------------

-----cp...@googlegroups.com wrote: -----To: cp2k <cp...@googlegroups.com>
From: Jörg Saßmannshausen
Sent by: cp...@googlegroups.com
Date: 04/21/2017 01:17PM
Subject: [CP2K:8961] calculation of a x-ray structure

[attachment "signature.asc" removed by Jürg Hutter/at/UZH]

[Jörg Saßmannshausen]

Hi Juerg,

thanks for the feedback.

I have tried what you suggested and I failed miserably.

I got this subsys section for the coordinates which are provided by the CIF
file with ABC and ALPHA_BETA_GAMMA taken from the CIF file:

&SUBSYS
&CELL
ABC [angstrom] 10.0980 16.7278 20.6290
ALPHA_BETA_GAMMA 93.744 94.745 91.931
PERIODIC xyz
&END CELL
&TOPOLOGY
CONNECTIVITY OFF
COORDINATE CIF
COORD_FILE_NAME 6d.cif
&END TOPOLOGY


However, I get that error message which puzzles me:

1 x
?

*******************************************************************************
* ___
*
* / \
*
* [ABORT]
*
* \___/ Syntax error in function string: Missing operator
*
* |
*
* O/|
1 x
?
*
* /| |
1 x
?
*
* / \ common/fparser.F:406
*
*******************************************************************************


===== Routine Calling Stack =====

5 read_coordinate_cif
4 coordinate_control_READ_COORDINATE
3 coordinate_control
2 topology_control
1 CP2K

1 x
?


I played around with PERIODIC (NONE and XYZ) in vain.

I don't get that error when I am using a xyz file.

Am I right to assume that for periodic calculations the POISSON_SOLVER needs
to be set to the default, i.e. PERIODIC?

For me it looks like the problem has to do with the CIF file but I am unable to
deduct the problem. I can open the same CIF file with Mercury for example.

Could somebody help me please?

All the best from a cold London

Jörg

[Matt W]

Hi Jörg,

yes that error comes during reading the cif file. I'm not sure how robust the CP2K cif parser is, the format can be fairly complex I think and maybe with different standards.

Can you save the (full, not just symmetry inequivalent) coordinates as xyz or pdb and use those - if you just want things to work for now?

Matt

[Jörg Saßmannshausen]

Hi Matt,

thanks for the quick reqply. I thought of doing that as well but I was not
quite sure what else, other than the coordinates, are being read from the CIF
file. I am usually doing single molecules hence I am sticking more to what
people advice me to do, rather than exploring.

So, is there anything else I need to bear in mind? In particular, am I right

to assume that for periodic calculations the POISSON_SOLVER needs to be set to
the default, i.e. PERIODIC?

Thanks again.

All the best from a sunny London

Jörg

> > <javascript:>

> >
> > > Winterthurerstrasse 190
> > > CH-8057 Zürich, Switzerland
> > > ---------------------------------------------------------------
> > >

> > > -----cp...@googlegroups.com <javascript:> wrote: -----To: cp2k <
> >
> > cp...@googlegroups.com <javascript:>>

> > email: j.sassma...@ucl.ac.uk <javascript:>

[Matt W]

Hi,

On Saturday, April 29, 2017 at 8:37:49 AM UTC+1, sassy wrote:

Hi Matt,

thanks for the quick reqply. I thought of doing that as well but I was not
quite sure what else, other than the coordinates, are being read from the CIF
file. I am usually doing single molecules hence I am sticking more to what
people advice me to do, rather than exploring.

Well maybe save the structure as pdb (or indeed re-export a new cif) from mercury or gdis or other visualization code you have available?

 

So, is there anything else I need to bear in mind? In particular, am I right
to assume that for periodic calculations the POISSON_SOLVER needs to be set to
the default, i.e.  PERIODIC?

Shouldn't be anything in particular, most cp2k defaults are set up for periodic systems. POISSON_SOLVER should be PERIODIC.

Matt

[Jörg Saßmannshausen]

Hi Matt,

thanks for the feedback.
I was actually using Mercury to save the file as xyz coordinates and that is
running fine now. Sorry, I thought I had mentioned that in my last email.

Now, what I am after are crystal packing effects. Does my current approach
takes these effects in considerations or would it be better to use a supercell,
i.e. expand the cell into all three dimensions?

I guess that was the original question I was after.

All the best from London

Jörg

[Matt W]

Hi again,

I'm a little confused, but if you take the periodic structure that comes from the XRD, and run it using the default periodic setup  specifying the appropriate unit cell, you'll have the crystal packing effects included.

Matt