The ABINIT Users Mailing List ( CLOSED )

["Andrew M. Rappe" <rappe@sas.upenn.edu>]

Dear Nichols,

Incorporating semicore states is an important advance.  However, it is 
not principally due to Martins, as far as I know.  There is a 1990 
paper by Teter and Allan that shows the importance of semicore for Ti:  
D. C. Allan and M. P. Teter, J. Am. Ceram. Soc. 73, 3247-50 (1990).

Many people (including us) do this routinely.

For one example, see our results for Ca (accurate energies for 3s, 3p, 
3d, and 4s) published in 1999:  N. J. Ramer and A. M. Rappe, Phys. Rev. 
B 59, 12471-8 (1999).

Perhaps the main importance of the Reis et al. paper is enabling 
semicore WITHIN the TM method.

OPIUM already permits you to make semicore psps (with optimized method) 
and has had this capability for years.

Your suggestion to fully implement TM for comparison purposes is 
already in progress.
Regards,
Andrew

On Oct 21, 2004, at 11:04 AM, Nichols A. Romero wrote:

> Hi Andrew,
>
> I will give it a try and let you know how it goes.
>
> I think that it would be great to have support for the TM PP. It would
> be great to have a direct comparison between the TM PP and optimized PP
> method. You are probably also aware of this. J.L. Martins published 
> made
> a significant towards explicitly incorporating semi-core states when
> there are multiple principal quantum numbers that have the same angular
> momentum channel.
>
> PRB 68, 155111 (2003)
>
> Opium could be the first PP program that incorporates this new
> advancement in PP theory.
>
> Bests,
> On Wed, 2004-10-20 at 23:56, Andrew M. Rappe wrote:
> > Dear Nichols,
> >
> > I would like to invite you to try the OPIUM code for your transition
> > metal pseudopotentials.
> > opium.sourceforge.net
> > In general, the optimized pseudopotential method will give better
> > convergence than TM.
> > In addition, the designed nonlocal pseudopotential method can
> > significantly improve the transferability.
> > (We also plan to add support for TM to the OPIUM code soon if that is
> > of great interest.)
> >
> > Sincerely,
> > Andrew M. Rappe
> >
> > On Oct 21, 2004, at 12:48 AM, Nichols A. Romero wrote:
> >
> > > Hi,
> > >
> > > I was wondering if anyone out there had a program that could convert
> > > the
> > > PP output format from the JL Martins ATOM code into one of the ABINIT
> > > formats. I would generate the PP with fhi98PP but it actually seems 
> > > to
> > > fail, complaining about convergence.
> > >
> > > Does anyone know if there are any significant differences in how the
> > > Hamann procedure for unbound states is implemented between these 
> > > codes?
> > > In my case, it is failing for 5f of Iridium in fhi98pp but it 
> > > succeeds
> > > ATOM code.
> > >
> > > Bests,
> > > --
> > > Nichols A. Romero
> > >
> > > 4-133 ESB, MC 704                           508 East Michigan Avenue
> > > 1110 West Green Street                      Apartment 12
> > > Urbana, IL 61801-3080                       Urbana, IL 61801-5176
> > > (217) 333-6108                              (217) 417-5210
> --
> Nichols A. Romero
>
> 4-133 ESB, MC 704                           508 East Michigan Avenue
> 1110 West Green Street                      Apartment 12
> Urbana, IL 61801-3080                       Urbana, IL 61801-5176
> (217) 333-6108                              (217) 417-5210