[CP2K-user] The strange fraction of the studied molecules in metadynamic simulation

[Marcella Iannuzzi]

Dear Tlyer,

The metadynamics settings in your input indicate that two distances have to 
be activated and explored. 
This can lead to bond breaking. How fast and whether the free energy 
surface is properly explored along this process depend on the other 
metadynamics settings, that I cannot judge not knowing specifically your 
problem and your goals.
 However, there are two aspects that you should consider. 
By breaking bonds it is a good idea to run spin polarised calculations, 
because of the possible formation of radicals. 
Actually, looking at the posted  trajectory,  I would say that one water 
molecule gets charged and start wandering hectically around.
Another remark is on the used walls that, for both collective variables, 
limit the maximum distance to 4 Å. 
This can significantly restrict the allowed space of configurations, 
unnaturally biasing the dynamics, which then becomes of difficult 
interpretation. 

Best
Marcella 


On Friday, March 26, 2021 at 3:33:06 AM UTC+1 ts... at gmail.com wrote:

> Dear users:
> Here is my input file and ouptput structure. At the begining of the 
> simulation, the AIMD fell reasonable, but As the simulation time 
> increases，The molecular structure has broken down strangely. I don't know 
> if it was an error in my input file. and maybe this is a normal results in 
> metadynamic simuilation. if you have time to give me a advice, i would 
> appreciate it.
> Best regards
> Tlyer
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20210326/f3072ebe/attachment.htm>