[QE-users] about the option "assume_isolated=2D"

[sohier]

Dear Juliana,

In principle, with the flag "assume_isolated= 2D", there should be 
nothing to take out from the band energies. That is, the vacuum level is 
zero. The flag is designed to give the result one would get with the 3D 
code in the limit of infinite distance between the periodic images.

A way to get the same number with the 3D code is, as you do, to subtract 
the vacuum level. The fact that 2D band energies are equivalent to the 
3D ones minus the vacuum level has been tested on a quite large number 
of 2D materials by some people in my group. It works well, except when 
there are out-of-plane polarization or other peculiarities in the 
electrostatics, which should not be your case.

In your case, given the 90 Ang you use, you should be relatively close 
to the limit of infinite distance between periodic image limit, at least 
within energies much smaller than 1eV! In fact, I would expect that you 
get the same band energies within some meV with or without the flag...

So... There is definitely a problem here. Could you send input, output, 
version of the code, etc... so I could have a closer look?

Thanks,
Thibault Sohier
THEOS, EPFL, Lausanne


On 19/02/2019 12:00, users-request at lists.quantum-espresso.org wrote:
> Date: Mon, 18 Feb 2019 14:58:10 +0100
> From: Juliana Morbec<jmmorbec at gmail.com>
> To:users at lists.quantum-espresso.org
> Subject: [QE-users] about the option "assume_isolated=2D"
> Message-ID:
> 	<CAFfGdaO2S_fRFxynk+mZjye_TdQf3zVEnQitDpeCcEVtdYebOw at mail.gmail.com>
> Content-Type: text/plain; charset="utf-8"
>
> Dear all.
>
> I have been trying to compute the band edge of 2D TMDCs and I have tested
> the option with and without "assume_isolated=2D". I noticed that this flag
> strongly affects the positions of the band edges. For example, the VBM/CBM
> of MoS2 change from -5.93/-4.20 eV with "assume_isolated=2D" to -4.96/-3.23
> eV without "assume_isolated=2D"; all values were computed with respect to
> the vacuum level, which is 0 eV in the first case and 0.08 eV in the second
> case; I used a quite significant vacuum area in the calculatons (~90 Ang).
>
> I read the paper "Density functional perturbation theory for gated
> two-dimensional heterostructures: Theoretical developments and application
> to flexural phonons in graphene" by  Sohier, Calandra add Mauri (Physical
> Review B 96, 75448), and I understood that the new method would make
> differences if one wants to investigate charged 2D materials, e.g. with
> charged defect, or 2D materials in a perpendicular electric field, but not
> in the case of pure MoS2, no charge, no field. I do not understand why such
> a large difference in the case of pure 2D layers.
>
> I will appreciate if someone could comment on this.
>
> Thank you for your time.
>
> Best regards,
>
> Juliana Morbec
> -- Juliana Morbec, PhD Research Associate - Prof. Kratzer's group 
> University of Duisburg-Essen, Germany https://jmmorbec.wordpress.com/

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20190219/543431e4/attachment.html>