can abinit do an all-electron calculation?

Total energy, geometry optimization, DFT+U, spin....

Moderator: bguster

Locked
onion2440
Posts: 34
Joined: Sat Sep 05, 2015 10:04 am

can abinit do an all-electron calculation?

Post by onion2440 » Mon Dec 03, 2018 4:09 am

Dear all,
Now I encounter a problem when I try to find the magnetic ground state of a material. However, The PAW PP of Abinit cannot show me the right one compare with the experiment. Since the all-electron method like Wien2k always have better performance in studying the magnetic properties, here I wanna know if there is a way to do an all-electron like calculation with Abinit? Or there is a way to improve the performance of the Abinit in studying the magnetic properties?

any helps will welcome!
Thank you.

onion2440
Posts: 34
Joined: Sat Sep 05, 2015 10:04 am

Re: can abinit do an all-electron calculation?

Post by onion2440 » Mon Dec 03, 2018 7:06 am

By the way, I have tested the energy cutoff, the kpt. And since I need to use GGA+U, I have to use the PAW method.

thank everyone for yours helps.

ebousquet
Posts: 469
Joined: Tue Apr 19, 2011 11:13 am
Location: University of Liege, Belgium

Re: can abinit do an all-electron calculation?

Post by ebousquet » Wed Dec 05, 2018 10:03 am

Dear onion2440,
If you don't give more details about your system, it'll be hard to reply to you.
You could have to initialize the spinat to a much larger value than the target one, you can also increase nline and increase nnsclo to 2 (thought this will make much slower calculations, then you can set that for the first few iterations).
All the best,
Eric

onion2440
Posts: 34
Joined: Sat Sep 05, 2015 10:04 am

Re: can abinit do an all-electron calculation?

Post by onion2440 » Thu Dec 06, 2018 3:39 am

ebousquet wrote:Dear onion2440,
If you don't give more details about your system, it'll be hard to reply to you.
You could have to initialize the spinat to a much larger value than the target one, you can also increase nline and increase nnsclo to 2 (thought this will make much slower calculations, then you can set that for the first few iterations).
All the best,
Eric


sorry for reply you later.
I will try it with your suggestions and will show you the results if the calculations are done.
The system is perovskite with an iron atom in it. Maybe you know that correctly predict the magnetic moments for iron is difficult for DFT codes, it's a hard work for me.

Thank you four help.
Chao Yang.

onion2440
Posts: 34
Joined: Sat Sep 05, 2015 10:04 am

Re: can abinit do an all-electron calculation?

Post by onion2440 » Mon Dec 10, 2018 1:48 am

Dear Eric,
I have tried your suggestions, I set a very large initio magnet moment, and also increase nline and increase nnsclo to 2.Then do the scf calculation with Abinit. However, the results are still not comparable with the Exp.
Do you know the magnetic ground state problem of HCP ion under high pressure? It's very similar with that.

Any way, thank you for your suggestions.
Best,
Chao Yang

ebousquet
Posts: 469
Joined: Tue Apr 19, 2011 11:13 am
Location: University of Liege, Belgium

Re: can abinit do an all-electron calculation?

Post by ebousquet » Fri Dec 14, 2018 4:07 pm

Dear Chao Yang,
Did you use the DFT+U to correct the d-electron self-interaction of Fe?
Which perovskite is it? I'm doing magnetic perovskites and so far it was OK for magnetism, thought the amplitude of the moment is often slightly smaller than the experimental one.
Best wishes,
Eric
Last edited by ebousquet on Mon Dec 17, 2018 4:48 pm, edited 1 time in total.

onion2440
Posts: 34
Joined: Sat Sep 05, 2015 10:04 am

Re: can abinit do an all-electron calculation?

Post by onion2440 » Mon Dec 17, 2018 5:28 am

Dear Eric,
I'm working on the magnetic perovskite with others. I'm not sure if it is problem to show it out. I'm very sorry for that. If they have any ideas ,I will let you know. By the way, I have used the DFT+U to correct the d-electron self-interaction.

Thank you,
Chao Yang

Locked