[CP2K-user] [CP2K:16412] Re: Large value of IASD

[Marcella Iannuzzi]

Hi .. 

The integrated spin density can be different from zero also if the 
multiplicity is 1, because it is 
\int_\Omega (\rho_{up}(r) - \rho_{down}(r)) dr
and not 
[\int_\Omega (\rho_{up}(r))dr] - [\int_\Omega (\rho_{down}(r))dr]
which instead is zero for multiplicity equal 1. 

Difficult to guess why the optimisation does not converge with the 
available information. 
It seems to be a magnetic system, which might be tricky to converge. 

Regards
Marcella


On Tuesday, December 28, 2021 at 9:05:31 AM UTC+1 gayathr... at gmail.com 
wrote:

>   Dear all,
>
> I have performed a  UKS calculation with *charge set to 0 and 
> multiplicity =1* (even  number of electrons in the system). I have made 
> around 15 attempts with UKS False, smearing on, different minimizes, 
> preconditioners and all of them failed to converge the SCF. 
>
> In one of those trials, the SCF struggled to converge and converged after 
> many steps. I terminated the simulation forcefully and got the following in 
> the output even for the converged SCF steps.
>
> Integrated absolute spin density  :                              59.
> 3323737862 <(332)%20373-7862>
>   Ideal and single determinant S**2 :                    0.000000      
> 29.954091
>
> For multiplicity set to 1, the values are unrealistic indicating severe 
> spin contamination and the results seem wrong.  I have attached the input 
> file. Can someone please help me in mitigating this issue?
>
> I will be very thankful for your help.
>

-- 
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+unsubscribe at googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/0c404ab3-3c57-4ed4-a96e-d259df620dd5n%40googlegroups.com.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20211228/53d5d35e/attachment.htm>