parameter of BSE

[zchen]

Dear Sir:
I have several questions
(1)After calculating GW correction I use -V qp to get the BSE parameters. I inserted the QP energy through KfnQPdb="E<./SAVE/db.QP" to the BS calculation. However, -V qp also listed many other parameters such as KfnQP_N, KfnQP_E,KfnQP_w, KfnQP_Z, XfnQPdb, XfnQP_N XfnQP_E, XfnQP_W XfnQP_Z etc. Can anyone briefly explain the physics meaning of those parameters? At moment I just use the default value. I am afraid to change any parameter.
(2) How can I insert GW energy to the RPA calculation. I used same method, -V qp. I cannot find parameter KfnQPdb to
add the QP energy. If I insert QP energy to the XfnQPdb="E<./SAVE/db.QP". The RPA spectrum did not change.
(3) We can find the indepencent particle approximation of the spectra from BSE and RPA output. I think two independent particle approximations should be same. However, they are different.

[myrta gruning]

Dear Chen

KfnQP_N, KfnQP_E,KfnQP_w, KfnQP_Z, XfnQPdb, XfnQP_N XfnQP_E, XfnQP_W XfnQP_Z etc. Can anyone briefly explain the physics meaning of those parameters? At moment I just use the default value. I am afraid to change any parameter.

First note that XfnQPdb, XfnQP_N XfnQP_E, XfnQP_W XfnQP_Z are the analogue of KfnQPdb, KfnQP_N KfnQP_E, KfnQP_W KfnQP_Z
K and X at the beginning are saying to you to which quantity the corrections are applied, either the kernel (K) or the Chi (X)

KfnQP_E and XfnQP_E are the
QP corrections using a scissor operator and valence/conduction stretching.
in the format
scissor | cond. stretching | val. stretching

KfnQP_N and XfnQP_N
is the number of neighbors you want to use in case you calculated the QP corrections on a different k mesh

Sincerely I never used the other variables, that I think regard the QP correction to W and changing the renormalization factor, I think that they are used in some particular case and the default values (that means that those corrections are not used) are OK. Maybe others can detail more on the use of these variables

2) How can I insert GW energy to the RPA calculation. I used same method, -V qp. I cannot find parameter KfnQPdb to
add the QP energy. If I insert QP energy to the XfnQPdb="E<./SAVE/db.QP". The RPA spectrum did not change.

As I said above KfnQPdb regard the kernel, that is why it is not in RPA where you have instead the same quantities for X (Chi)
Are you getting any warning message in the log or report?

(3) We can find the indepencent particle approximation of the spectra from BSE and RPA output. I think two independent particle approximations should be same. However, they are different.

Are you sure you are doing the same calculations, using the same parameters?
In general if you encounter any problem you should post inputs and outputs of the calculation in question otherwise it is very difficult for us to understand what is going on

Finally, please fill-in your signature, so that it will appear automatically in each post.

I hope this helps,

Myrta

[zchen]

Dr. Myrta:
Thank you.

Zhifan chen
Dept of Physics
Clark Atlanta University
USA