Atoms are not relaxing, but successful optimization

Matt W MattWa... at
Fri Mar 23 10:23:30 UTC 2012

On Thursday, March 22, 2012 1:17:07 PM UTC, Samuel Lamphier wrote:
> Thank you for your response, I used crystal maker to derive a 2x2x2 
>  supercell, I exported the orthogonal coordinates directly. However, I 
>  thought that if I incorporated a vacancy that I would see a much 
>  larger relaxation as I do using classical simulations. The change was 
>  on the order of 0.000001 for cp2k and during classical simulations, I 
>  observed changes of 0.1 to 1.5A during relaxation. 
Sorry, didn't notice that you had included an oxygen vacancy. In that case 
without seeing more of your output it is impossible to say what occurs.

Does the wavefunction converge?  Does CP2K report geometry optimization 
steps in the output? Perhaps you should quickly check you can relax a water 
molecule or similar just to be sure the problem is with this input.

BTW, I suspect that the Ce basis set and pseudo potential have never been 
used in the wild, so be very careful. 

PS Does anyone fancy optimizing a GGA pseudo for Ce?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <>

More information about the CP2K-user mailing list