XAS input

[marci]

Well, as I said, in my experience 20÷30 eV above the homo is
the interesting energy region for NEXAS.
It is difficult to say to how many unoccupied states this corresponds,
as it is system-dependent.
Example: for ice, 128 H2O molecules and 640 occupied orbitals,
using a  6-311Gxx basis set, 400 ADDED_MOS cover about 12 eV

I use to run the ground state without ADDED_MOS and then restart
from the optimized MOS to compute XAS all the time I like, using
always the same RESTART.
You can also restart a XAS calculation, for example to increase the
number of ADDED_MOS.

ciao
Marcella

On Oct 30, 10:28 am, "carlo antonio pignedoli" <c.pig... at gmail.com>
wrote:
> Ciao Marcella,
> that's more or less all I needed, thanks a lot!!
> In addition.. if I am interested only in the core level shift in a
> bulk calculation,
> what would be your suggestion for a reasonable number of  ADDED_MOS
> (depending on the
> size of the system)?
>
> Thanks again
>
> Carlo
>
> On Thu, Oct 30, 2008 at 10:14 AM, marci <marcella.ia... at psi.ch> wrote:
>
> > Hi Carlo,
>
> > when running core level spectroscopy simulations, you should specify
> > from which core state the electron is extracted. This is done
> > by giving the index of the atom on which the core state is located and
> > the
> > type of the state, for example 1s.
> > Then the corresponding molecular orbital has to be identified.
> > The search is limited to the lowest n MOS, as given by the input
> > keyword STATE_SEARCH.
> > Typically, first a localization procedure is run only over these n MOS
> > and then the
> > location and character of each rotated MOS is verified.
> > Once the desired state has been identified, its occupation number is
> > changed and a new SCF is started,
> > using  the ground state MOS as initial guess.
>
> > For the emission spectrum (XES) only the occupied valence orbitals are
> > needed. Instead,
> > a certain number of unoccupied states should be added to obtain a
> > reasonable
> > absorption spectrum.
> > The implemented method is suitable for near-edge-XAS, i.e. within a
> > range of about 20÷30 eV above the homo.
> > This is why the  ADDED_MOS keyword is present in the XAS section.
> > The number you give indicates how many states are added to the initial
> > set of MOS that has been used for the ground state calculation.
> > It is duplicated because in principle you do not need many unoccupied
> > states for the ground state calculation.
> > If you already added unoccupied states for the ground state
> > calculation, say 10, the number of states given in XAS%ADDED_MOS, say
> > 100,
> > is added to the first 10, i.e. the new SCF with the excited core state
> > is run with 110 unoccupied states.
>
> > I hope this clarifies your doubts,
>
> > ciao
> > Marcella
>
> > On Oct 30, 9:36 am, "c.pignedoli" <c.pig... at gmail.com> wrote:
> >> Dear all,
> >> I'm not so sure to fully understand the input cards of the XAS
> >> section.
>
> >> In particular, I don't know the meaning of
>
> >> STATE_SEARCH  n
>
> >> and the meaning of ADDED_MOS if another card ADDED_MOS is present in
> >> the SCF section
>
> >> Can anybody help me?
>
> >> Ciao
>
> >> Carlo