[CP2K-user] Optimizing stretched Al(111) monolayer

[Luca]

Dear Marcella,
thank you for the kindly response. I am trying to reproduce the 
Al(111)/Al2O3 reported  in this work.
DOI: 10.1103/PhysRevLett.103.076103.
I was not able to reproduce it mainly because of the "clusterizzation" of 
Al(111) on top of the surface.
Therefore, I simply built a prototype 
(i.e.  Al monolayer ) trying to find a set-up able to bias the result way. 
The aim of this tentative with Al monolayer  was to understand how "to 
improve" 
the setup the input to reproduce the above structure, if it possible.

Luca 

Il giorno mercoledì 18 novembre 2020 alle 12:01:55 UTC+1 Marcella Iannuzzi 
ha scritto:

> Dear Luca,
>
> I cannot see how a stretched single layer of Al atoms could show any 
> metallic character. 
> I am wondering whether the previous results that you mentioned are to be 
> trusted. 
> Best
> Marcella
>
>
>
> On Tuesday, November 17, 2020 at 7:18:56 PM UTC+1 Luca wrote:
>
>> Dear all,
>> I am trying to optimize a single layer of Al (111) in the (XY free) plane 
>> by constraining the Z component. Al (111) is stretched a = ~ 3 Ang.
>> Using a pure PW + PBE scheme, the system converges towards a lattice as 
>> reported in the attached file Al_111.pdb
>> Using cp2k I cannot reproduce the same result because the system 
>> reorganizes itself into smaller metal domains, breaking the original Al 
>> (111) lattice.
>> I tried with several XCs, but I could not reproduce the "metallic state" 
>> obtained as in the PW scheme, which is necessary to treat a more complex 
>> system.
>> Therefore, I don't really care about the correct behavior of this simple 
>> system, I want try to find a way to reproduce PW results using cp2k.
>> Do you have any suggestions on how to bias the result of cp2k towards the 
>> result obtained with PW-PBE scheme?
>>
>> Thank you,
>>
>> Luca
>>
>>
>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20201118/5f9a6b73/attachment.htm>