The ABINIT Users Mailing List ( CLOSED )

[Anurag Chaudhry <anuragchaudhry@yahoo.co.uk>]

Hi,

Thanks for your reply.

I have tried specifying the occupation by hand (same as what you referred to as occopt 0) for an inorganic insulator doped with Cerium (Ce 3+).  In Ce 3+ we have only one 4f electron so I specified occopt 0 with LDA to get the excited state (no spin polarization in this case).

but with other lathanide series elements like Eu2+ I need spin polarization and with that I am not sure how to do a constrainted LSDA since for the ground state Abinit returns occupation states like 0.85 ... near the fermi level.  how to take  these fractional numbers and get the excited state ?

as an example this is what I get when I do spin polarized calculations for the occupation
              1.000000  1.000000  1.000000  1.000000  1.000000  1.000000
              1.000000  1.000000  1.000000  1.000000  1.000000  1.000000
              1.000000  1.000000  1.000000  1.000000  1.000000  1.000000
              1.000000  1.000000  1.000000  1.000000  1.000000  1.000000
              1.000000  1.000000  1.000000  1.000000  1.000000  1.000000
              1.000000  1.000000  1.000000  1.000000  1.000000  1.000000
              1.000000  1.000000  1.000000  1.000000  1.000000  1.000000
              1.000000  1.000000  1.000000  0.999999  0.999999  0.999999
              0.999999  0.999999  0.999999  0.999999  0.999999  0.999998
              0.999998  0.999998  0.999998  0.999996  0.999996  0.999996
              0.999995  0.999987  0.999987  0..999987  0.999987  0.999979
              0.999978  0.999978  0.999977  0.999950  0.999948  0.999946
              0.999946  0.999944  0.999943  0.999942  0.999940  0.999819
              0.999817  0.999812  0.999806  0.999753  0.999751  0.999745
              0.999736  0.002378  0.002313  0.002126  0.001895  0.000624
              0.000603  0.000546  0.000445  0.000139  0.000137  0.000134
              .........

and for another system with Ce3+ I get occupation states as
              1.000000  1.000000  1.000000  1.000000  1.000000  1.000000
              1.000000  1.000000  1.000000  1.000000  1.000000  1.000000
              1.000000  1.000000  1.000000  1.000000  1.000000  1.000000
              1.000000  1.000000  1.000000  1.000000  1.000000  1.000000
              1.000000  1.000000  1.000000  1.000000  1.000000  1.000000
              1..000000  1.000000  1.000000  1.000000  1.000000  1..000000
              0.000000  0.000000  0.000000  0.000000  0.000000  0.000000
              0.000000  0.000000  0.000000  0.000000  0.000000  0.000000
              0.000000  0.000000  0.000000  0.000000  0.000000  0.000000

thanks,
Anurag

----- Original Message ----
From: Anglade Pierre-Matthieu <anglade@gmail.com>
To: forum@abinit.org
Sent: Sunday, 4 May, 2008 11:56:22 PM
Subject: Re: [abinit-forum] constrainted LSDA with Abinit

Hello,

I think it is possible : have you tried to use "occopt 0" and specify
occupation numbers by hand ("occ") in your input file ?

regards

PMA

On Mon, May 5, 2008 at 6:04 AM, Anurag Chaudhry
<anuragchaudhry@yahoo.co.uk> wrote:
>
> Hello,
>
> I would like to know if it is possible to do constrained LSDA calculation
> with the possibility to explicitly specify the occupation of an excited
> state.  For a ground state calculation with spin polarization Abinit gives
> non-integer band occupation for the two spins so I am not sure how to
> specify the occupation of the excited state.
>
> Thank you very much for your time and help.  I am still learning so please
> pardon me if my question is not clear, I will try to formulate it again.
>
> regards,
> Anurag
>
>  ________________________________
>  Sent from Yahoo! Mail.
>  A Smarter Email.



--
Pierre-Matthieu Anglade

---

Sent from Yahoo! Mail.
A Smarter Email.