[CP2K-user] Geometry optimisation of organic molecule on fcc Au(111) slab

[Marcella Iannuzzi]

I think that the cell dimensions in your input are wrong, because they do 
not correspond to an integer number of replica of the Au(111) slab.
I also think that your slab cannot be a good model because it has only 2 
layers. In may opinion it makes no sense to freeze all slab's atoms.
Moreover charged systems on metals are not well defined with DFT and 
periodic boundary conditions, and it is very risky to try this type of 
calculations if you are not very sure of what you are doing and how to 
interpret the results. 
Unless you want to model a Au cluster interacting with a molecule, you need 
periodic boundary conditions. 

Regards
Marcella



On Monday, November 16, 2020 at 2:48:54 AM UTC+1 dre... at gmail.com wrote:

> Thank you for reply!
>
> I expected organic molecule moves more from it's initial position. 
> Otherwise it's almost same as I built in avogadro software. My ultimate 
> goal is to perform some geometry optimization with applied electric field 
> or potential on z axis. I would like to imitate charge surface and compare 
> some structural orientation changes to uncharged system. 
>
> Could you please advise me smth in this sense. 
>
>  
>
> Regarding my cell dimensions, z is twice larger then x and y. Does it 
> sound right?
>
>  
>
> Do you think while optimization I should keep system non periodic with 
> poison wavelet  or periodic in xy or xyz with poison analytic?
>
>  
>
> Thank you,
> Dmitrii    
>
> On Sunday, November 15, 2020 at 11:32:43 PM UTC+11 Marcella Iannuzzi wrote:
>
>>
>>
>> Dear Dmitrii
>>
>> The lateral dimension of a slab cannot be approximately large enough, but 
>> it has to be a multiple of the proper 2D lattice constant.
>> Moreover, 2 layers of a metallic slab are for sure not sufficient to 
>> reproduce the right electronic properties. For Au(111) the minimum would be 
>> 4 layers.
>> This said, which type of motion do you expect? All Au atoms are frozen, 
>> and the molecule interact only weakly via dispersion forces. 
>> Regards
>> Marcella
>>
>>
>> On Sunday, November 15, 2020 at 11:00:26 AM UTC+1 dre... at gmail.com 
>> wrote:
>>
>>> Good day, I run geometry optimisation of organic molecule on fcc Au(111) 
>>> slab with CP2K QS. Cutoff of 800 Ry and real_cutoff 70 Ry, smearing on, 
>>> dispersion correction with DFTD3 PBE, and poison  periodic xyz  with 
>>> wavelet solver. My cell dimension is 15.00000 15.00000 26.00000, whereas 
>>> size of Au slab is about 12*12*4 A. I froze all atoms atoms in list, and 
>>> allowed only organic molecule to be equilibrated with GC optimiser. My 
>>> system converged and completed optimisation, however my system did not move 
>>> much. I am concerned that I did smth wrong with setting for convergence or 
>>> periodicity. Could you please have a look at my files and give me feedback.
>>>
>>>  
>>>
>>> Thank you,
>>>
>>> Dmitrii
>>>
>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20201116/a7728e34/attachment.htm>