[CPMD-list] Help on obtaining a buckled Si(001) surface?

[Juerg Hutter]

Hi

a few observations regarding your input:
- the optimized surface (most likely) will have lower
  symmetry than you initial clean surface. By using
  POINT GROUP AUTO you fix the symmetry.
- Silicon is one of the few elements where normconserving
  PP can be used with cutoffs as low as 12 Ry. There is no
  need to use USPPs and a cutoff of 30 Ry.
- NLCC is not needed either.

regards

Juerg

----------------------------------------------------------
Juerg Hutter                   Phone : ++41 1 635 4491
Physical Chemistry Institute   FAX   : ++41 1 635 6838
University of Zurich           E-mail: hutter at pci.unizh.ch
Winterthurerstrasse 190
CH-8057 Zurich, Switzerland
----------------------------------------------------------


On Sat, 12 Apr 2003, Du Aijun wrote:

> Dear all,
>
> Recently, I performed a geometry optimization from a clean si(001)
> unreconstruced surface to obtain a bulckled dimer surface. The convergence
> is very slow and it seems the geometry don't changed too much after 20
> geometry steps. Here is my imput file. I hope someone can check it for me!
> Any comments and suggestions will be highly appreciated!
>
> &CPMD
>     OPTIMIZE GEOMETRY
>     HESSIAN UNIT
>     LBFGS
>     MEMORY BIG
>     PCG MINIMIZE
>     CONVERGENCE
>      1.D-5  5.D-5
>     MAXSTEP
>      500
>     STRUCTURE BONDS ANGLES
>  &END
>
>  &SYSTEM
>    ANGSTROM
>    POINT GROUP
>     AUTO
>    SYMMETRY
>     6
>    CELL
>     15.358868    1.0    1.3   0.0 0.0 0.0
>    CUTOFF
>     30.0
>    STATES
>    128
>    TESR
>     1 1 1
>  &END
>
>  &ATOMS
> *014-Si-gpw-n-campos.uspp BINARY NEWF NLCC
>    LMAX=D
>    64
>  -6.7205 -6.7205  5.4310
> ...
> ...
> ...
> ...
>   2.8802  4.8004  9.5043
>   6.7205  4.8004  9.5043
> CONSTRAINTS
> FIX ATOMS
> 16
> 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
> END CONSTRAINTS
>  &END
>  &DFT
>    NEWCODE
>    FUNCTIONAL GGA
>    GC-CUTOFF
>    1.D-5
>  &END
>
>
>
>
>
>
>
>
>
> _________________________________________________________________
> Ãâ·ÑÏÂÔØ MSN Explorer:   http://explorer.msn.com/lccn/
>
> _______________________________________________
> CPMD-list mailing list
> CPMD-list at cpmd.org
> http://www.cpmd.org/mailman/listinfo/cpmd-list
>