Page 1 of 1

Rare earth

Posted: Fri Jan 02, 2009 6:07 am
by Kitaura
Dear vasp users,

I'm working on band structure calculation of compounds containing rare earth metals.
I have a question on PPs treating f-orbital as a core.

When I calculate, for example Gd compounds possessing 4f7 electronic configuration with PPs treating f-orbital as a core, should I perform the calculation using ISPIN=2 (spin polarized system) ? I think if I use PPs treating f-orbital as valence, I should definitely set ISPIN =2. But in this case, what should I do ?

any comments and advices are greatly appreciated.

Thanks in advance

Rare earth

Posted: Thu Jan 08, 2009 3:07 pm
by admin
This of course depends on the compound and the valence electron configuration of its atoms. Staying with your Gd example: even if you use the PP which puts the f-electrons into tthe core, you have 1 unpaired 5d electron. --> in this case, please use ISPIN=2

Rare earth

Posted: Mon Feb 16, 2009 7:11 pm
by wangfei
Hi admin,

What if I use Gd_3 in the calculation of a compound such as Gd5Ge4? Since the f electrons are in the core in Gd_3, do I still have to use ISPIN=2 for the unpaired 5d electron?

Rare earth

Posted: Tue Feb 17, 2009 11:19 am
by admin
yes, of course

Rare earth

Posted: Thu Feb 19, 2009 5:39 pm
by wangfei
Thank you for the reply, admin!
I still have one more question about this unpaired 5d electron. In a compound, this electron can be paired with the valence elecron from another atom (for my case, with a 4p electron from Ge). So if I use ISPIN=1, it still makes sense. Am I right?
<span class='smallblacktext'>[ Edited Thu Feb 19 2009, 06:46PM ]</span>