Top Page > Browsing
Is restart needed for the final state calculation for binding energies of core levels?
Date: 2020/07/06 23:57
Name: Pavel Ondracka   <pavel.ondracka@email.cz>

Dear all,

When I run the TiC216-CH3.dat (final state) input, without the restart files from the TiC216.dat (initial state) present I end with a calculation which reaches convergence, but has a total energy of -10488.752 instead of the expected -10489.553 in http://www.openmx-square.org/openmx_man3.9/node193.html

The manual implies that the restart keyword is optional for the final state calculations, but I can't get the correct results when I start from scratch... Do I need to tweak the mixing somehow so it can find the correct value even when starting from scratch? Or what is the actual issue?

Best regards
Pavel
メンテ
Page: [1]

Re: Is restart needed for the final state calculation for binding energies of core levels? ( No.1 )
Date: 2020/07/09 22:43
Name: T. Ozaki

Hi,

If you do not provide the restart file, you need to switch "scf.coulomb.cutoff" off.
This is because rho_f is zero resulting in zero V^P and and V^NP is only taken into account.
The long range Coulombic interaction is totally ingored due to the Coulombic cutoff technique.

I performed the following two calculations, and obtained the expected results.


With the following keywords:

scf.restart on
scf.restart.filename TiC216
scf.coulomb.cutoff on
scf.core.hole on

I got -10489.553359996646 as total energy.

With the following keywords:

scf.restart off
scf.restart.filename TiC216
scf.coulomb.cutoff off
scf.core.hole on

I got -10489.553980187460 as total energy.


Regards,

TO
メンテ
Re: Is restart needed for the final state calculation for binding energies of core levels? ( No.2 )
Date: 2020/07/10 23:13
Name: Pavel Ondracka  <pavel.ondracka@email.cz>

Thanks, it works. Just so I understand correctly, this (scf.coulomb.cutoff off) should be done also in the case when I do core-hole calculation (without the restart files from the initial state calculations) with charged system (core hole in non-metalic systems and/or molecules)?
メンテ
Re: Is restart needed for the final state calculation for binding energies of core levels? ( No.3 )
Date: 2020/07/11 08:40
Name: T. Ozaki

Hi,

As for charged bulks, a restart file has to be provided, and "scf.coulomb.cutoff" needs to be switched on as

scf.restart on
scf.restart.filename ***
scf.coulomb.cutoff on
scf.core.hole on

As for charged molecules, the long range Coulombic interaction has to be cut, leading to the following keywords:

scf.restart off
scf.coulomb.cutoff on
scf.core.hole on

Regards,

TO
メンテ
Re: Is restart needed for the final state calculation for binding energies of core levels? ( No.4 )
Date: 2020/09/15 15:23
Name: Pavel Ondracka  <pavel.ondracka@email.cz>

To follow up on the previous questions, I'm now trying to model a polymer interaction with a metallic surface. I can't really model the polymer, so I'm in fact using just a dimer (molecule with ~30 atoms, the surface has around 500). I want to compare binding energies in the molecule away from the surface and than with a second model with the molecule at the surface (after pushing the molecule to the surface and some MD run) as well as changes in the metal atoms near to the interaction site. The question is what scheme should I use to calculate the binding energies? I'm considering to using the charged molecule scheme for the calculation when the molecule is away from the surface and than for the charged bulk when at the surface. Or should I just use a separate model with the molecule only for the initial BE calculations, instead of using the starting model for the MD runs with the surface present?

I don't need an absolute binding energies, I'm thinking about using C1s binding energy of some carbon atom in the molecule not directly interacting with the surface as a reference level to calculate the relative shifts of carbons and oxygens at the interaction site (and compared to the non-interacting molecule). This should work OK, right?

Also does it make sense to use the Effective screening medium method in this settings? Will it even work with the binding energies calculations?
メンテ

Page: [1]

Thread Title (must) Move the thread to the top
Your Name (must)
E-Mail (must)
URL
Password (used in modification of the submitted text)
Comment (must)

   Save Cookie