[Wien] regarding mixing parameter in hybrid calculations

[tran at theochem.tuwien.ac.at]

Hi,

First, the standard value is 0.25 and not 0.025.

For the dielectric constant, there was alreday some explanations here:
https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg14936.html
https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg14875.html

alpha_opt at V=V_exp means that the dielectric function (and alpha_opt)
is calculated only at V_exp and then used for all other volumes.
See also this:
https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg15531.html

No, it is not mandatory to do PBE calculation before YS-PBE0. Starting
YS-PBE0 from PBE electron density (case.clmsum) may just help to reduce
the number of SCF iterations with YS-PBE0.

Of course it is possible to optimize lattice parameters with YS-PBE0.
You just need to add the "-hf" option in the optimize.job file.
What is not possible with YS-PBE0 is to optimize the position of
atoms in the unit cell (because the forces are not implemented for
YS-PBE0).

FT


On Saturday 2018-08-04 14:46, Dr. K. C. Bhamu wrote:

>Date: Sat, 4 Aug 2018 14:46:02
>From: Dr. K. C. Bhamu <kcbhamu85 at gmail.com>
>Reply-To: A Mailing list for WIEN2k users <wien at zeus.theochem.tuwien.ac.at>
>To: A Mailing list for WIEN2k users <Wien at zeus.theochem.tuwien.ac.at>
>Subject: [Wien] regarding mixing parameter in hybrid calculations
>
>Dear Dr. Tran,
>
>In the recent past, I did some calculations using YS-PBE0 for MgO and results were well reproduced (wien2k Exercise).
>
>I just read your paper (J. Phys.: Condens. Matter 25 (2013)) and get to know that we can control the mixing parameter using various ways.
>
>1) Standard value (0.025)
>2) By fitting  (alpa_opt)
>3) alpa_opt (V=V_exp)
>
>I have queries for option 2 and 3 alpa:
>
>2) If I am not wrong (after having a detailed search), the value of dielectric constant used to get alpa_opt is the value of dielectric constant from
>Rel(epsilon) at energy 0.013610 eV, right?
>
>3) How to get alpa_opt at V-V_exp?
>
>
>
>
>
>My next query is:
>
>It is reported in the Wien2k exercise that to do YS-PBE0 calculations we need to do a simple PBE scf.  So if we really need to perform PBE calculations before
>going to YS-PBE0, I am wondering how to use YS-PBE0 to optimize the lattice parameters and to get structural properties?
>
>
>Could you please advice me on my concerns?
>
>* in above text "_" means an underscore.
>
>
>regards
>
>
>
>