spin-polarized calculation with biased spin: how to obtain ground state of oxygen atom

[Gabriele Tocci]

Dear CP2K users,

I apologise if a similar post has already been discussed but I only found one where a similar matter has been pointed out, that is http://groups.google.com/group/cp2k/browse_thread/thread/f1f4d594212348c1/c7f8b5c2f5a64789?lnk=gst&q=BS&pli=1 , but this did not help.

I am trying to compute the total energy of a single oxygen atom doing a spin-polarized energy calculation, however the Mulliken population analysis printed out in the end of the output file  gives a total magnetic moment is 0, with both spin-up and spin down states having an occupation of 3 in the 2s and 2p orbitals, whereas one spin-state should be 4 and the other one 2, giving rise to a spin moment of 2.

I have tried to fix this in the same way as explained in the post above, by using the section BS and biasing alpha and beta electrons in the chosen spin states but this didn`t fix it. Is this the correct way to do it but I am missing something or is there another way?

Below is my relevant input file:

  &SUBSYS
    &CELL
ABC       10      15       20
    &END CELL
    &COORD
       SCALED
O 0.5 0.5 0.5
&END COORD
    &KIND O
     BASIS_SET DZVP-MOLOPT-GTH-q6
      POTENTIAL GTH-PBE-q6
     &BS
        &ALPHA
          NEL     2
          L       1
          N       2
        &END
        &BETA
          NEL    -2
          L       1
          N       2
        &END
     &END
    &END KIND
  &END SUBSYS


Here there are the relevant parts of the output where first at the beginning the occupation has changed according to what specified, but at the end, after the total energy has been calculated, the mulliken population analysis shows again an occupation of 3 for both spin up and spin down states. This can ultimately result in wrong total energies, right?

Atomic guess: The first density matrix is obtained in terms of atomic orbitals
               and electronic configurations assigned to each atomic kind

 Spin 1

 Guess for atomic kind: O

 Electronic structure
    Total number of core electrons                                          2.00
    Total number of valence electrons                                       8.00
    Total number of electrons                                              10.00
    Multiplicity                                                   not specified
    S   [  2.00] 2.00
    P      6.00

After some lines...

Spin 2

 Guess for atomic kind: O

 Electronic structure
    Total number of core electrons                                          2.00
    Total number of valence electrons                                       4.00
    Total number of electrons                                               6.00
    Multiplicity                                                   not specified
    S   [  2.00] 2.00
    P      2.00



But finally, at the end of the calculation the wrong occupancies appear

MULLIKEN POPULATION ANALYSIS

 # Atom  Element  Kind  Atomic population (alpha,beta)   Net charge  Spin moment
      1     O        1           3.000000     3.000000    -0.000000    -0.000000
 # Total charge and spin         3.000000     3.000000    -0.000000    -0.000000


Any help would be kindly appreciated.
Thanks.

[Valerio Bellini]

Dear Gabriele,
you have to specify the MULTIPLICITY label in your input file
otherwise the program assume you have S=0 (multiplicity 1, for an even
number of electrons)
(
http://cp2k.berlios.de/manual/CP2K_INPUT/FORCE_EVAL/DFT.html#desc_MULTIPLICITY
)
cheers,
Valerio

[Valerio Bellini]

ops,
you wanted a singlet..
but if you do DFT you should say to the program it is a spin polarized
calculation
adding 'LSD T'.
Did you set it?
cheers,
Valerio

[Valerio Bellini]

sorry for the sequence of e-mail..
I read it wrong, you wanted a S=1.
So it holds my first message,
as a matter of facts your output say " Multiplicity not specified".
cheers,
Valerio

[Gabrielet=16]

Dear Valerio,

Thank you very much. The multiplicity tag fixed it! I had already used
the LSD tag for a spin-polarized calculation, but I had to switch off
BS.

Kind regards,
Gabriele