Beyond the Plasmon-Pole approximation in LiF inout database

[kopinjol]

Hi everyone!

I was going through the tutorials and I just realized that the input database for the tutorial "Beyond the Plasmon-Pole approximation in LiF" doesn't exist. I can probably create the PwScf outputs. But It would be an immense help if I could somehow get the yambo input files.

Thanks in advance,
kopinjol baishya

[Daniele Varsano]

Dear kopinjol,
that's quite strange. In the tutorial page:
clicking on the YAMBO CORE DATABASES and REFERENCE files download
you should be able to download the databases. I tried and it looks to work.
Anyway this is the link
to download the files. Let us know if you experience problems. About yambo inputs, you can find them in the links
browsing the tutorial.

Cheers,
Daniele

[kopinjol]

Hi Daniele,

Thanks for the quick reply, really appreciate it. I used your link to download the databases, but it contains only the Solid Al files. The Input folder has only one input file named 01_lifetime, which is for the Solid Al part of the tutorial. The input files for the LiF part of the tutorial are not there. I am here talking about the second tutorial "Beyond the Plasmon-Pole approximation: real-axis GW and lifetimes" and the second part if this tutorial " Beyond the Plasmon-Pole approximation in LiF ". I also looked at the databases from other parts of the tutorial page, i.e. from " Solid LiF: excitons at work ". But none of them contain the input files for the LiF part of the second tutorial. I am sorry if i am missing something obvious.

Thanks,
kopinjol

[Daniele Varsano]

Dear Kopinjol,
if you go through the tutorial, for instance here:
[08] QPs in real-axis GW (Newton solver): 09_QP_Real_Axis (yambo -g n )

You can click on "09_QP_Real_Axis" and the input file will appear, I tried just now
and it looks to work.

About the databases: you can download the LiF, form the previous tutorial: fantastic dimension, clicking on:
YAMBO CORE DATABASES and REFERENCE files (suggested) (106Mb): download.

Hope this helps,

Cheers,

Daniele

[kopinjol]

Hello Daniele,

Thanks a lot for your prompt reply. I will try out what you suggested just now.

Thanks again,
kopinjol

[kopinjol]

Hi Daniele,

I tried what you suggested. I clicked on 08_QP_PPA and copied the contents of the input file that appeared. This is how it looked:

gw0 # (R)(GW) GoWo Quasiparticle energy levels
ppa # (R)(Xp) Plasmon Pole Approximation
HF_and_locXC # (R)(XX) Hartree-Fock Self-energy and Vxc
em1d # (R)(Xd) Dynamical Inverse Dielectric Matrix
EXXRLvcs RL # (XX) Exchange RL components
%QpntsRXp
1|19| # (Xp) Transferred momenta
%
%BndsRnXp
1|20| # (Xp) Polarization function bands
%
NGsBlkXp RL # (Xp) Response block size
%LongDrXp
1.000000|0.000000|0.000000| cc # (Xp)(Xp) Electric Field
%
PPAPntXp eV # (Xp) PPA imaginary energy
%GbndRnge
1|50| # (GW)(G) W bands range
%
GDamping eV # (GW)(G) W damping
dScStep eV # (GW) Energy step to evalute Z factors
DysSolver # (GW) Dyson Equation solver (`n`,`s`,`g`)
ExtendOut # (F)(GW) Print all variables in the output file
%QPkrange
1|1|3|6| #
%
%QPerange
1|1|0.0|-1.0| #
%



But when I tried running Yambo with it (yambo -F Inputs/08_QP_PPA -J PPA), I end up getting an error message. The message reads:

<---> [01] Files & I/O Directories
<---> [02] CORE Variables Setup
<---> [02.01] Unit cells
<---> [02.02] Symmetries
<---> [02.03] RL shells
<---> [02.04] K-grid lattice
<---> [02.05] Energies [ev] & Occupations
<---> [03] Transferred momenta grid
<---> [04] Bare local and non-local Exchange-Correlation
<---> [05] Dynamic Dielectric Matrix (PPA)
<---> [FFT-X] Mesh size: 24 24 24
<---> [WF-X loader] Wfs (re)loading |####################| [100%] --(E) --(X)
<---> [X-CG] R(p) Tot o/o(of R) : 222 6144 100
<---> [X] Upper matrix triangle filled
<01s> Xo@q[1] 1-2 |####################| [100%] --(E) --(X)
<01s> X @q[1] 1-2 | | [000%] --(E) --(X) ** On entry to ZGETRF parameter number 4 had an illegal value

This is actually the whole output. The last line is the error. It seems it causes an error in caliing some lapack routine. Do you know what might be going wrong? I had noticed some of the variables that are not assigned any value in the input file, do not have any default values. Could that be the problem?

Thanks in advance,
kopinjol

[Daniele Varsano]

Dear Kopjnol,

first of all, be careful on the choice of the input file:
you copied the 08_QP_PPA, that means a Quasi particle calculation in PPA (plasmon pole approximation).
If you want to go beyond the PPA you should copy the 09_QP_Real_Axis .
Anyway, the problem you encountered as you argued is that the input file is incomplete:
the input file from the tutorial tell you what are the fundamental variable to be set, but some
of them have to be filled by hand.

I suggest you to generate by yourself the input file by line command:
see here the different option depending on the kind of calculation you want to perform.In this way you will have the default value assigned, but *careful*, does not mean that the values
you will have in the input are rhe right value, as many variables have to be set according to convergence test you have to perform.

For instance for QP calculation in plasmon pole approximation:
yambo -g n -p p

for a QP real axis calculation:
yambo -g n -c

Check here some useful combination.

In particular in your input the values for the following variable are missing:
EXXRLvcs RL # (XX) Exchange RL components
NGsBlkXp RL # (Xp) Response block size
PPAPntXp eV # (Xp) PPA imaginary energy
GDamping eV # (GW)(G) W damping
dScStep eV # (GW) Energy step to evalute Z factors
DysSolver # (GW) Dyson Equation solver (`n`,`s`,`g`)
ExtendOut # (F)(GW) Print all variables in the output file


In any case, let me suggest you to take your time and go through the tutorial, for instance the basic GW you find just above the one you are doing, or may start from the LiF you find in the fantastic dimension tutorial, in order to get familiar with the yambo use.

Cheers,

Daniele

[kopinjol]

Hello Daniele,

Thanks for the reply and the explanation. I really appreciate it. I have already gone through all the tutorials. I am going to look at the links you sent and try to make the input file by hand.

Here I would like to make a quick observation. I was repeating some calculations on the SiH4 molecule using a real space code, Parsec and RGWBS. I didn't have to use supercell as this is a real soace code and ideal for finite systems. The bse spectra i get from this code is blue shifted from the yambo spectra. I have played around with a lot of input variable, such as boundary sphere radius, grid spacing and number of bands. But I still get a blue shifted spectrum consistently.

Has anyone had similar results? I was wondering why this blue shifted result occurs. Do you think there are factors other than quantum confinement?

Thanks again,
kopinjol

[Daniele Varsano]

Dear Kopinjol,
the blueshift you observe does not surprise me too much. As you can see in the tutorial,
the agreement with the experiment of the tutorial calculation is not at all excellent.
When comparing the tutorial results with literatures, or other codes,
it is important to have in mind that the tutorial we prepared are meant for yambo schools,
and in doing that the parameter of the calculations in the input files are not set for having
converged results, but for running in reasonable time with few cpus, and for taking confidence
with theory and code.
So the message we always gives, beside the examples, that the calculations for production run need convergence, convergence and convergence.
Beside that there is the fact that when using plane waves code for isolated system, vacuum effect plays an important
role and have to be carefully checked. In order to eliminate interaction with repeated images, in yambo
there are developed coulomb cutoff technique (see Rozzi, Varsano, Marini et al. PHYSICAL REVIEW B 73, 205119 2006).
In order to use that have a look at the Documentation, and the argument is extensively discussed in the forum.
The vacuum effects plays a role both in the GW calculation and in BSE spectra. If you do not remove this spurious effects,
you usually get an underestimation of the gap and this could be the reason of the blueshift you observe.
Beside that, SiH4 it is not at all a simple system: see for instance Rohlfing, Louie, PHYSICAL REVIEW B 62, 4927 (2000), where
difficulties in GW calculation for this system are highlighted. Anyway, what is the amount of the observed blueshift? This is an information
that could be useful for other users reading the forum. Thanks for noticing it.
Cheers,

Daniele

[kopinjol]

Hello Daniele,

I really appreciate your quick reply. Sorry I couldn't reply right away. I had some teaching responsibilities to take care of. Thanks a lot for the detailed explanation. I checked out the papers.

But my question was slightly different I think. I am aware of the yambo spectra displaying a blue-shift as compared to the experiments. But I was comparing the yambo spectra with a real space (not plane-wave) code. I was using Parsec and RGWBS to be precise. And the BSE results I get from RGWBS are blue shifted as compared to the yambo spectra. The blue shift is about 1 eV. I was wondering if you had any comments about the result. RGWBS could be giving the blue shift for quantum confinement effects. But my boundary sphere radius is large enough I think (about 24 aus).

Thanks again for the prompt and detailed reply. I really appreciate it.

Thanks,
kopinjol