[CPMD-list] PP

[GU ZHIHUI]

 Dear all,

Thank you very much for your information!! It is very helpful!

When I generate the Pseudopotential of Hydrogen by ATOMC, The
Pseudo-wavefunction does not match the All-Electron wavefunction which is
created by ATOMC too beyond the R cut-off( in the range of 0.3, 0.4 ...0.8).
The following is the sample of input file. 

&ATOM
    atom=H
    core=0
    valence=1
    lda=BLYP
    gcx = BECKE
    gcc = LYP
    states
    1    S   1.0
    MARTINS
WAVEFUNCTIONS
 channels 1
 0  0.50  0.0
&END

Could you please tell me what problem is? Should I change the reference
state? For example,

 1  s  0.8
 2  s  0.2


By the way, what is the unit of R in wavefunction?


Many thanks,

Zhihui GU


-----Original Message-----
From: cpmd-list-bounces at cpmd.org [mailto:cpmd-list-bounces at cpmd.org] On
Behalf Of Axel Kohlmeyer
Sent: Thursday, July 20, 2006 5:18 PM
To: CPMD
Subject: Re: [CPMD-list] PP

On 7/20/06, GU ZHIHUI <zhihui.gu at chemail.tamu.edu> wrote:

since ari already covered the first part of your question, here a comment to
the second part:

> By the way, Could anyone gives me some suggestions about the R cutoff 
> for valence orbitals and R core cutoff??

finding good cutoff values for the various channels in pseudopotentials is
part of the 'black magic' behind creating psp-generation. there are a few
rules that you have to follow (e.g. r_c must be large enough, that there is
no node in the corresponding all-electron wavefunction for all r > r_c), but
then there are also several 'should be' rules (e.g. the r_c for different
channels should be very similar) that can help to avoid ghost states and
finally you have the property, that smaller r_c may increase transferability
at the cost of needing a higher plane wave energy cutoff.

for a good introduction into creating norm-conserving pseudopotentials and
the associated pitfalls, i would recommend looking at the fhi98pp code and
the associated tutorial. see:
http://www.fhi-berlin.mpg.de/th/fhi98md/fhi98PP/index.html
what is stated here applies to CPMD as (in fact you can even convert the
fhi98pp generated pseudopotentials to CPMD format). the atomc code is
lacking most of the pp-testing instrumentation, so you have to test those
potentials directly with CPMD.

regards,
    axel.

> Many thanks!!
>
>
> Zhihui Gu
> _______________________________________________
> CPMD-list mailing list
> CPMD-list at cpmd.org
> http://cpmd.org/mailman/listinfo/cpmd-list
>
>
>


--
=======================================================================
Axel Kohlmeyer   akohlmey at cmm.chem.upenn.edu   http://www.cmm.upenn.edu
  Center for Molecular Modeling   --   University of Pennsylvania
Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
=======================================================================
If you make something idiot-proof, the universe creates a better idiot.
_______________________________________________
CPMD-list mailing list
CPMD-list at cpmd.org
http://cpmd.org/mailman/listinfo/cpmd-list