Periodic boundary conditions

Matt W mattwa... at
Tue Jun 26 10:07:42 UTC 2018

CP2K does not wrap molecules back into the box when writing an MD 
trajectory - so you can see the 'diffusion' of a molecule. This can be 
useful for analysis.

Whether you expect molecules to diffuse that far depends on the system / 
length of simulation time. You could try running a small box of water with 
classical MD and see how that looks.

As you note the molecule _is_ wrapped back into the cell when all 
calculations are done. You can safely wrap back into the cell for 
visualization etc.


On Tuesday, June 26, 2018 at 10:58:22 AM UTC+1, Dawid das wrote:
> Hi, 
> I would like to repeat my question, please.
> Best wishes,
> Dawid Grabarek
> W dniu poniedziałek, 18 czerwca 2018 10:00:14 UTC+2 użytkownik Dawid das 
> napisał:
>> Dear CP2K Users,
>> I am doing QM/MM MD simulation using PBC only for MM part (the QM cell is 
>> not periodic). My cell is 
>> orthorombic and I am a bit surprised as in course of simulation some 
>> water molecules drift away to vacuum 
>> when I use VMD to visualize my trajectory.
>> Now, I am aware that if PBC work it does not matter as the image of this 
>> molecule "enters on the other side"
>> however this visualization makes me to ask a question whether I do my 
>> setup correctly.
>> So in course of MD what I get in XYZ trajectory file is not wrapped by 
>> CP2K and can I safely wrap those 
>> molecules/atoms later on with a script? In other words, does CP2K save 
>> only coordinates of molecules 
>> found in starting geometry but not its images as they leave the box?
>> Best wishes,
>> Dawid
>> PS This is part of my input that should simply apply PBC, am I right?
>>   &MM
>>     &POISSON
>>       &EWALD
>>         EWALD_TYPE SPME
>> !3.5/12 Angstrom
>>         ALPHA .292
>>         GMAX  32 32 16
>>         EWALD_ACCURACY 1.0E-6
>>       &END EWALD
>>   &END MM
>>   &SUBSYS
>>     &CELL
>>       ABC [nm] 7.60236   6.53052   6.03591
>>       ALPHA_BETA_GAMMA 90 90 90
>>     &END CELL
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <>

More information about the CP2K-user mailing list