[QE-users] Large difference between experimental lattice constant and the relaxed one w/ and w/o dft-d3

[Mladenovic Marko]

Dear Houcine,


First of all, which exchange-correlation functional are you using? If you are using LDA, that may explain the underestimation of the lattice constants. Taking the experimental lattice and relaxing the structure without relaxing the cell ('relax' instead of 'vc-relax') is a valid approach. I personally prefer to relax both, provided that the level of theory I am using gives the lattice constants that are accurate enough. The system doesn't look to me like something to which I would directly apply DFT-D3, but I haven't studied this type of materials.


Best wishes,

Marko

________________________________
From: users <users-bounces at lists.quantum-espresso.org> on behalf of 후신 부지드 <bouzid at skku.edu>
Sent: Friday, December 20, 2019 2:57:28 PM
To: users at lists.quantum-espresso.org
Subject: [QE-users] Large difference between experimental lattice constant and the relaxed one w/ and w/o dft-d3


Dear QE users,

I'm working on a layered material (SnS2) slightly doped with transition metals (TM)

the experimental c lattice constant of this Trigonal system is around 6.88 A as measured by XRD (slightly less with Mn doping)

DFT calculation without any dispersion force correction gives 6.49 A

by including a dft-d3 correction during the relaxation it is even less ~ 5.95

- Any advice what value should I consider ? It is correct to take the experimental value without any relaxation ?

- Same inquiry about London dispersion force correction, is it needed to use it in this case for next steps (scf and nscf caculations) since it is giving a big difference compared with the experimental value ?

Any help is appreciated!

Best Regards


Houcine BOUZID
Sungkyunkwan University,
Korea.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20191220/d3c265db/attachment.html>