[CPMD-list] qery on KS energy calculation and eigenvalues

[Ari.P.Seitsonen at iki.fi]

Dear Dr Menon,

> 1)
> In Kohn-Sham Energy calculation, how to calculate "number of empty states
> and number of occupied states" for a molecule?. In CPMD manual, in the
> example for one water molecule, in one case this value is 'zero' and other
> case is 'four'. Could anyone explain this?
>  
> KOHN-SHAM ENERGIES
>      N
>  
> N = ?? 

This depends on how many unoccupied/virtual states you want to obtain.
N=0 means that the true Kohn-Sham states (i.e. those which diagonalise
the Lambda matrix on the right hand side of the KS equations) are
wanted, without any unoccupied states; this corresponds to a
diagonalisation of the matrix <\phi_j|H|\phi_i>.

> 2)
> Is it possible to do geometry optimization in Kohn-Sham energy calculations?
> I mean, is it possible to do calculations with
> OPTIMIZE GEOMETRY
> KOHN-SHAM ENERGIES
>  
> The reason why I asked this is because it says both are mutually exclusive.
> But in the example section (pg 53) it is given that both optimization step
> and Kohn-Sham Energies can be used. 
> OPTIMIZE WAVEFUNCTION
> KOHN-SHAM ENERGIES
>  
> When I tried to do that particular example, the output says "both are
> mutually exclusive" and stops.
> So, Is it is possible to do geometry optimization in Kohn-Sham energy
> calculation, then what modification I have to do?

It is not possible to do them in the same run, this wouldn't bring
much advantage: the Kohn-Sham states (neither KS wavefunctions nor
eigenvalues) are not needed during the geometry optimisation. You can
calculate the Kohn-Sham states in another job with a RESTART.

> 3)
> I also tried to calculate the Eigenvalues by printing the keyword "PRINT ON
> INFO EIGENVALUES". My input file looks like this:
> &CPMD
>     OPTIMIZE WAVEFUNCTION
>     INITIALIZE WAVEFUNCTION RANDOM
>     STRUCTURE BONDS ANGLES
>     PRINT ON INFO EIGENVALUES
>     STORE
>     2
>     ISOLATED MOLECULE
> &END
>  
> But with this input, eigenvalues are neither printed in the output file nor
> in any other files. If I want the eigen values in the output file, what I
> have to do?

I believe that the PRINT INFO ... is more meant for dynamical runs. In
jobs including only the optimisation of the electronic structure one
can use the Kohn-Sham energies.

    Greetings,

       apsi

-- 
-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-
 Ari Paavo Seitsonen / Ari.P.Seitsonen at iki.fi / http://www.iki.fi/~apsi/
 Tel +41 1 635 44 97 / Fax +41 1 635 68 38 / GSM +41 79 719 09 35      
 Anschrift: Physikalisch Chemisches Institut (PCI), Universität Zürich (UniZh)
 Indirizzo: Winterthurerstraße 190, CH-8057 Zürich
 Address:   Schweiz / Svizzera / Suisse / Svizra / Switzerland