The ABINIT Users Mailing List ( CLOSED )

["Chun Li " <chun.li@mf.mpg.de>]

Hi, Prasenjit,
Thank you very much. 
How about the first question? That is, how to choose the proper
pseudopotential for certain calculation?

Chun Li

-----Original Message-----
From: Prasenjit Sen [mailto:prasen@mri.ernet.in] 
Sent: 2004年7月6日 11:18
To: forum@abinit.org
Subject: Re: [abinit-forum] computation of the total energy


On Tue, 6 Jul 2004, Chun Li  wrote:

> Dear Abinit users,
>
> I just tried making the computation of the total energy for ZnO
> according to the tutorial. I have two questions:
>
> 1)       I choose the Troullier-Martins pseudopotentials for both Zn
and
> O atoms, is it reasonable? And if I want to perform response-function
> calculation later, will the pseudopotentials properly work?
> 2)       When I set the nband to 4 or 8, the log file indicates that
> there are not enough bands to get charge balance right. It is until I
> set the nband to 12 that the code can execute normally (when I set it
to
> 10, it also can perform normally with the values of fcart all being
> zero). What's the reason then?
> Thanks a lot.

  Dear Chun,
       in TM psp, Zn has 12 valence electrons and O has 6. So in ZnO
with
one Zn and one O atoms, there are 18 electrons. Therefore, you need at
least 9 bands to hold all the electrons. However, typically you should
have a few empty bands as well for better convergence.

   Prasenjit Sen

 =======================================================================
  Harish-Chandra Research Institute     Phone: +91-532-2667509 Extn.
2007
  Chhatnag Road, Jhunsi                  Or    +91-532-2667511*2007
  Allahabad, UP 211019
       INDIA                            Fax:   +91-532-2568036
 =======================================================================