[Wien] Cu(111) Surface State

[Chiung-Yuan Lin]

Dear Prof Blaha,
  Thanks for reminding about the thickness effect.
However, the case I am actually interested in is
to latter put dilute impurities on this surface,
and I think even a 5- or 7-layer slab can describe
the surface-state wavefunction near the impurities
fairly well (though become less accurate deeper into
the slab or vaccum). Do you agree?
  If WIEN can obtain these image potential states
accurately, how can I construct an appropriate slab?
Does (21-layer slab) + (8-layer vaccum) sounds OK?
Are there any other parameters I should tune besides
the thickness?

Chiung-Yuan

On Wed, 10 Mar 2004, Peter Blaha wrote:

> If you are interested on surface states, you MUST use a very thick slab
> with lots of vacuum. Forget your 5 and 7-layer calculations. In these
> models the 2 surfaces of the slab interact with each other and this leads
> to a large splitting of the two surface states (there are 2 surface
> states, because in our slab model we have two surfaces!) This interaction
> can be via the slab or via the vacuum, so both parameters must be checked.
>
> Your results are fine, when you see two (nearly) degenerate
> bands, which are both the surface state, in your bandstructure plot.
>
> Some shift in the position of the eigenvalues with respect to experiment
> is "normal". This is the well known DFT problem.
>
> PS: WIEN can obtain these image potential states, but your slab must be
> appropriate. We can even quantitativel explain the splitting of the
> surface state into two states in Au due to spin-orbit interactions, but
> for your Cu this is irrelevant.
>
>  > Dear all,
> >   I was doing slab calculation for Cu(111). From experiments, there is
> > a surface state 0.39eV below the Fermi energy. I did 5-layer, 7-layer,
> > and 21-layer slabs (with 4, 8, and 8 layers of vaccum respectively).
> > The calculated surface states are 0.3, 0.44, and 0.54 eV respectively.
> > The charge convergence criterions are 0.01, 0.01, and 0.005 respectively.
> > All 3 cases are calculated using the potential option PBE96. I am
> > wondering if those calculated energies are the best Wien2k can do.
> >   A previous question was discussed between Stolbov and Prof Blaha
> > (in the "new digest", 2003 December, "Plotting surface potential").
> > >From that It seems PBE96 is OK for Cu(111) surface potential.
> >   On the other hand, on Phys Phys. Rev. B 195110-4 (2001) it says that
> > wien97 cannot yield the asymptotic image potential so that the Cu(111)
> > surface state turns out to be incorrect.
> >   I am thinking if the suggestion of the above PRB paper applies to my
> > case. Or, the image-potential problem has been fixed in Wien2k, the
> > surface-state energies I calculated are off the experiment due to other
> > technical problems.
> >   I appreciate very much if anyone can give me suggestions.
> >
> > Chiung-Yuan
> >
> > _______________________________________________
> > Wien mailing list
> > Wien at zeus.theochem.tuwien.ac.at
> > http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> >
>
>
>                                       P.Blaha
> --------------------------------------------------------------------------
> Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
> Phone: +43-1-58801-15671             FAX: +43-1-58801-15698
> Email: blaha at theochem.tuwien.ac.at    WWW: http://info.tuwien.ac.at/theochem/
> --------------------------------------------------------------------------
>
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>