[Pw_forum] C pseudo regeneration

[Hande Ustunel]

Dear Andrea,

Many thanks. That fixed the problem. I had indeed intended it to be 1.3 but
had an error in my automatized script.

Best regards,
Hande

On Mon, 27 Aug 2007, Dal Corso Andrea wrote:

> On Mon, 2007-08-27 at 15:16 +0300, Hande Ustunel wrote:
> > Dear forum members,
> > 
> > I have a problem concerning the reproduction of one of the C
> > pseudopotentials. I'm trying to reproduce the C.pz-rrkjus.UPF
> > pseudopotential as best I can using the input provided below. Although I
> > get excellent agreement with the log derivatives and test cases (pls see
> > below), even with cutoffs of 40/400 Ryd, I get a diamond lattice constant
> > that is 1.2% too little and a rather bumpy E vs a curve. I've tried a
> > rather exhaustive set of different parameters but I haven't been able to
> > generate a smooth enough curve. The C.pz-rrkjus.UPF from the web site on
> > the other hand gives a very nice smooth curve with an error of only 0.9%
> > for the same calculation parameters. I've posted both curves here.
> > 
> >     http://www.physics.metu.edu.tr/~hande/curves
> > 
> > I guess the errors are not terribly different but should I be worried about
> > the bumpiness? If yes, it would be really helpful if anyone could spot the
> > difference between my input and the pseudo online. If not, any suggestion
> > on improving my pseudo would be more than welcome. Thank you very much in
> > advance. 
> > 
> > Best regards,
> > Hande
> > 
> > PS : I'm using version 3.2.
> > 
> > Input :
> > -------
> > 
> > &input
> >      atom='C',
> >      rlderiv=3.0,
> >      eminld=-4.0,
> >      emaxld= 4.0,
> >      deld  = 0.02,
> >      nld   = 3,
> >      config=' 1s2 2s2 2p2 3d0.00',
> >      iswitch=3,
> >      prefix='c'
> >      dft='pz',
> >  /
> >  &inputp
> >    pseudotype=3,
> >    lloc=2,
> >    tm=.false.,
> >    nlcc=.false.,
> >    file_pseudopw='C.pz.uspp-rrkj.UPF',
> >  /
> > 5
> > 2S  1  0   2     0.00  1.3 1.6
> > 2S  1  0    0     -0.4  1.3 1.6
> > 2P  2  1   2     0.00  1.6 1.6
> > 2P  2  1    0     -0.1  1.6 1.6
> 
> The core radii of the 2P wavefunctions seem strange to me. This is a
> norm conserving pseudopotential for the 2P functions. 1.3 1.6 usually
> work.
> 
> Andrea
> 
> 
> > 3D  3  2   -2.0   2.00   1.3    1.3
> >  &test
> >    nconf=1,
> >    configts(1)='2s2 2p2 ',
> >  /
> > 
> > 
> > 
> > Test results :
> > --------------
> > Atomic configuration : s2 p2
> >      n l     nl             e AE (Ry)        e PS (Ry)    De AE-PS (Ry)
> >      1 0     2S   1( 2.00)       -1.00195       -1.00210        0.00015
> >      2 1     2P   1( 2.00)       -0.39860       -0.39861        0.00002
> > 
> > Test case 1 : s2 p1.75
> >      n l     nl             e AE (Ry)        e PS (Ry)    De AE-PS (Ry)
> >      1 0     2S   1( 2.00)       -1.20025       -1.20038        0.00012
> >      2 1     2P   1( 1.75)       -0.58967       -0.58941       -0.00026
> > 
> > Test case 2 : s1.00 p1.50
> >      n l     nl             e AE (Ry)        e PS (Ry)    De AE-PS (Ry)
> >      1 0     2S   1( 1.00)       -2.38716       -2.38539       -0.00178
> >      2 1     2P   1( 1.50)       -1.76433       -1.76042       -0.00391
> > 
> > Test case 3 : s1.00 p1.50
> >      n l     nl             e AE (Ry)        e PS (Ry)    De AE-PS (Ry)
> >      1 0     2S   1( 1.00)       -2.38716       -2.38539       -0.00178
> >      2 1     2P   1( 1.50)       -1.76433       -1.76042       -0.00391
> > 
> > 
> > 
> 

-- 
Hande Ustunel
Department of Physics
Office 439
Middle East Technical University
Ankara 06531, Turkey
Tel : +90 312 210 3264
http://www.physics.metu.edu.tr/~hande