[CPMD-list] Help with molecular crystals -- atoms are too close

[Ewald Pauwels]

Hello,

I think the origin of your problem is that you have input 3 molecules in 
your unit cell instead of the 2 that were given in the X-ray structure. 
Remember that CPMD will periodically repeat the molecules in your unit 
cell in all directions. Hence the 'third' molecule is in effect 
redundant and perfectly overlapping with atoms of nearby virtual cells.

Sincerely,
Ewald.


-- 
Ewald Pauwels
Center for Molecular Modeling
Laboratory of Theoretical Physics
Ghent University
Proeftuinstraat 86
B-9000 Gent
Belgium
http://molmod.ugent.be
ewald.pauwels at UGent.be
+32 9 264 65 76







okan.esenturk at nist.gov wrote:

>Hi all,
>I am new on using CPMD.
>-SNIP-
>This input file is made from crystal data using Mercury. It has three full 
>molecules. Crystal unit only has two (Z=2) but I did not want to input partial 
>molecules. It is an orthorhombic unit with the given unit cell parameters. 
>
>I have not optimized the geometry or wavefunction. I'd like to calculate the 
>crystal phonon modes of the structure. Optimizing geometry will change the 
>distances and symmetry, I believe.
>
>I have also tried with standard potential comes with the CPMD but both fails 
>in the same place reporting atoms are too close. I've checked the given 
>coordinates (from the input file or from the output file) to make sure and all 
>the atoms are in proper distances from each other. 
>
>Would anybody know why I would receive such an error and the why it is 
>not properly running.  output file is given below.
>
>Thanks you very much.
>Okan Esenturk
>
>info: running on multi processor MPI-LAM system. Compiled for 64 bit 
>opteron and xeon systems. Using version CPMD 3.11.1.
>
>  
>




-- 
No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.1.407 / Virus Database: 268.12.12/461 - Release Date: 2/10/2006