Page 1 of 1

How it works the Quasi-Particle shift?

Posted: Mon Mar 17, 2014 1:32 pm
by javad
Dear developers

I'm trying to do a calculation of optics with a GW-shift,
I have done

yambo -o c -V QP

I got in the input

% XfnQP_E
0.000000 | 1.000000 | 1.000000 | # [EXTQP Xd] E parameters (c/v) eV|adim|adim
%

I understood that the first number is the correction to the gap, what are the other two numbers? and how can I use them?

with best regards
Javad Exirifard

Re: How it works the Quasi-Particle shift?

Posted: Mon Mar 17, 2014 1:53 pm
by Daniele Varsano
Dear Javad,
I'm trying to do a calculation of optics with a GW-shift,
I have done

yambo -o c -V QP
OK, but be careful, if you do a linear response calcualtion (RPA) applying a QP shift it is not totally consistent, the QP correction is meant for a BS calculation.
I got in the input

% XfnQP_E
0.000000 | 1.000000 | 1.000000 | # [EXTQP Xd] E parameters (c/v) eV|adim|adim
%

I understood that the first number is the correction to the gap, what are the other two numbers? and how can I use them?
The other two variable refers to the slope of the valence/conductions QP energies (coming from a GW calculations for isntances) with respect the KS energies. You can fit them by plotting E_qp vs E_lda. Sometimes beside the QP shift you have a different slope for valence and conduction energies. Considering

Code: Select all

QP_shift |1 |1 
you apply only the shift and it is the so called scissor operator approximation.
See here for a description:
If you cannot apply a simple shift (as for instances in many molecules), you need to calculate the QP energies you need and apply them using the Qpdb variable with the syntax:

Code: Select all

"E < example/(n)db.QP" 
here (n) is if you linked with the netcdf libraries.

Hope it helps,
Best,
Daniele