Orbitals calculation in MD-simulations

Question on input files/tags, interpreting output, etc.

Please check whether the answer to your question is given in the VASP online manual or has been discussed in this forum previously!

Moderators: Global Moderator, Moderator

Locked
Message
Author
Dragonling
Newbie
Newbie
Posts: 3
Joined: Tue Nov 12, 2019 10:44 am

Orbitals calculation in MD-simulations

#1 Post by Dragonling » Tue Mar 24, 2020 5:27 pm

Dear VASP-community,

I am studying the MD-simulations, but I have a big question about Bandgap calculations and HOMO and LUMO orbitals: how they are calculated?
As far as I understand, each atom position in MD can be respresented as a different scf-calculation cycle. Are orbitals energy averaged later? Or it just calculates in some other manner?
Also, if we will consider only 2 atoms, it is obvious that they can oscillate in phase or out of phase, and we will obtain different results for these cases.

merzuk.kaltak
Administrator
Administrator
Posts: 277
Joined: Mon Sep 24, 2018 9:39 am

Re: Orbitals calculation in MD-simulations

#2 Post by merzuk.kaltak » Mon Mar 30, 2020 9:28 am

The HOMO and LUMO energies depend on NELECT, the total number of (valence) electrons considered.
Orbital energies are not averaged during an MD. In fact, they are determined after each ionic step and written to the OUTCAR file.
Make sure that NBLOCK=1, so that the full information after each ionic step is written to file.
A convenient way to extract the orbital energies around the band gap from the OUTCAR is to use grep.
For instance, to extract two orbital energies below and above the HOMO in our tutorial example, would have to do:

Code: Select all

[./Si_Liquid_Melting]$ grep -A2 -B2 "   128  " OUTCAR
    126       5.1167      1.99995
    127       5.1167      1.99994
    128       5.1167      1.99994
    129       5.6976      0.00003
    130       5.6979      0.00003
--
    126       5.0859      1.99999
    127       5.1221      1.99994
    128       5.1677      1.99920
    129       5.6486      0.00056
    130       5.6708      0.00017
--
...
The first block corresponds to the first configuration in XDATCAR, the second block to the second configuration and so on.

Dragonling
Newbie
Newbie
Posts: 3
Joined: Tue Nov 12, 2019 10:44 am

Re: Orbitals calculation in MD-simulations

#3 Post by Dragonling » Mon Mar 30, 2020 11:04 am

Thank you very much for your reply!
I know how to extract HOMO and LUMO through grep command.
But the question is more about orbitals. Can you please tell me, how they are calculating during MD? What is the difference between orbitals calculation during MD and non-MD runs? And if no difference, should orbitals have the same energy in MD (with any temperature) and non-MD setup?

merzuk.kaltak
Administrator
Administrator
Posts: 277
Joined: Mon Sep 24, 2018 9:39 am

Re: Orbitals calculation in MD-simulations

#4 Post by merzuk.kaltak » Mon Mar 30, 2020 12:56 pm

The orbitals are represented in the projector augmented wave basis set as explained [url=wiki/index.php/PAW_method]on this website[url] and the references therein.
The chosen thermostat is chosen with [url=wiki/wiki/index.php/MDALGO]MDALGO[url] and affects the total energy. Please read the cited references.

Locked