[Wien] CXZ and P

[Peter Blaha]

> Firstly, the (117) P structure is really no any
> problem run 'x symmetry'. But if I use (117) CXZ
> structure and don't change gamma to 90.001 degree, it
> will give error  information 'Fortran Stop -3

Please update your WIEN code from the web.


> Secondly, When calculating (113) slab, we compared CXZ
> and P structure, and found their surface energy almost
> the same. So we thought CXZ could be instead of P

If you have checked it, that's ok. I have had cases were it
works perfectly well, but others where it does not produce
correct resutls and unfortunately I was not able to find this
problem so far. It was just a warning.


> time. But '7 EIGENVALUES BELOW THE ENERGY -8.00000'
> was occured during iteration. We checked case.scf,
> :DIS is converged very well no any vibration.

Probably approximate linear dependency for the s states!

a) Any reason to run with such small spheres ?
    ATOM  1  Pd1        ATOM  2  Pd2
 RMT(  1)=2.10000 AND RMT(  2)=2.10000
 SUMS TO 4.20000  LT.  NN-DIST= 5.27460

Setting RMT to 2.5 will make the calculations 10 times faster.

b) Why did you go up to RKMAX=8.4 ?
:RKM  : MATRIX SIZE 4626LOs: 117  RKM= 8.40  WEIGHT= 1.00  PGR:

Using APW+lo an RKMAX=7-8 should be enough. In particular ALWAYS do first the
calculations with a small RKMAX and check it later by increasing it.
For your surface you probably want to relax the forces. The forces give a
perfect test of RKMAX (do the relaxation with the smallest possible RLMAX,
as long as you get "identical (i.e. similar)" forces.

c) Switch to LAPW for s and p states and treat only d with APW+lo


> Finally, yes, we also noticed maybe 7 layers of (117)
> is not so relvant. 9 layers is also calculated.

The "7 layers" do NOT have a single Pd atom which is 12-fold coordinated !
Your "9-layer" slab at least has one. (I would definitely even go for
a thicker slab!!!)

It is completely irrelevant to run calculations with huge RKMAX, but then
the geometrical model is too limited.

When using large RMTs and reasonable RKMAX you can easily treat cells with
many more atoms!

                                      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/
--------------------------------------------------------------------------