BSE directly with the dielectric constant calculated in GW

[ljzhou86]

I want to do a BSE for 2D metallic silicene directly with the dielectric constant calculated in the GW calculation according to this post:viewtopic.php?f=13&t=436
Due to the use of the latest version of yambo.3.4.0, I generate the yambo.in by yambo -o b -k sex -y d -c -p p as followed:

optics # [R OPT] Optics
bse # [R BSE] Bethe Salpeter Equation.
bsk # [R BSK] Bethe Salpeter Equation kernel
bss # [R BSS] Bethe Salpeter Equation solver
ppa # [R Xp] Plasmon Pole Approximation
rim_cut # [R RIM CUT] Coulomb interaction
em1d # [R Xd] Dynamical Inverse Dielectric Matrix
em1s # [R Xs] Static Inverse Dielectric Matrix
RandQpts= 3000000 # [RIM] Number of random q-points in the BZ
RandGvec= 1 RL # [RIM] Coulomb interaction RS components
CUTGeo= "box z" # [CUT] Coulomb Cutoff geometry: box/cylinder/sphere
% CUTBox
0.00000 | 0.00000 | 12.00000 | # [CUT] [au] Box sides
%
CUTRadius= 0.000000 # [CUT] [au] Sphere/Cylinder radius
CUTCylLen= 0.000000 # [CUT] [au] Cylinder length
Chimod= "Hartree" # [X] IP/Hartree/ALDA/LRC/BSfxc
BSEmod= "causal" # [BSE] resonant/causal/coupling
BSKmod= "SEX" # [BSE] IP/Hartree/HF/ALDA/SEX
BSSmod= "d" # [BSS] (h)aydock/(d)iagonalization/(i)nversion/(t)ddft`
BSENGexx= 30249 RL # [BSK] Exchange components
BSENGBlk= 205 RL # [BSK] Screened interaction block size
#WehCpl # [BSK] eh interaction included also in coupling
% BEnRange
0.00000 | 10.00000 | eV # [BSS] Energy range
%
% BDmRange
0.10000 | 0.10000 | eV # [BSS] Damping range
%
BEnSteps= 100 # [BSS] Energy steps
% BLongDir
1.000000 | 0.000000 | 0.000000 | # [BSS] [cc] Electric Field
%
% BSEBands
1 | 20 | # [BSK] Bands range
%
% QpntsRXs
1 | 43 | # [Xs] Transferred momenta
%
% BndsRnXs
1 | 20 | # [Xs] Polarization function bands
%
NGsBlkXs= 205 RL # [Xs] Response block size
% LongDrXs
0.1000E-4 | 0.000 | 0.000 | # [Xs] [cc] Electric Field
%
% QpntsRXp
1 | 43 | # [Xp] Transferred momenta
%
% BndsRnXp
1 | 20 | # [Xp] Polarization function bands
%
NGsBlkXp= 1 RL # [Xp] Response block size
% LongDrXp
1.000000 | 0.000000 | 0.000000 | # [Xp] [cc] Electric Field
%
PPAPntXp= 27.21138 eV # [Xp] PPA imaginary energy

(1) Are there some errs in the yambo.in? And how to process smoothly? When I do such calculation, the job crashed with some err information like this:

<01m-30s> Xo@q[42] 1-2 | | [000%] --(E) --(X)
<01m-31s> P01: Xo@q[42] 1-2 |####################| [100%] 01s(E) 01s(X)
<01m-32s> X @q[42] 1-2 | | [000%] --(E) --(X)
<01m-32s> P01: X @q[42] 1-2 |####################| [100%] --(E) --(X)
<01m-32s> [X-CG] R(p) Tot o/o(of R) : 3256 11202 100
<01m-32s> Xo@q[43] 1-2 | | [000%] --(E) --(X)
<01m-33s> P01: Xo@q[43] 1-2 |####################| [100%] 01s(E) 01s(X)
<01m-33s> X @q[43] 1-2 | | [000%] --(E) --(X)
<01m-33s> P01: X @q[43] 1-2 |####################| [100%] --(E) --(X)
<01m-33s> [07.01] Main loop
<01m-33s> [WARNING]Fractional e/h occupations. Causal BSEmode forced.
<01m-33s> [WARNING] The system is a metal but Drude term not included.
<01m-33s> [08] Bethe-Salpeter Kernel
<01m-33s> [BSE] Kernel dimension : 2344
<01m-33s> [08.01] Screened interaction header I/O
<01m-33s> [WARNING]BS section skipped. PP/Em1s DB does not fit/exist
<01m-33s> [09] BSE solver(s)
<01m-33s> [10] Game Over & Game summary

(2) If I generate yambo.in by unadding -p p, this job can be done well. While, this "[WARNING]Fractional e/h occupations. Causal BSEmode forced." still existed, what's the reason. I note a reference: Wei, W.; Dai, Y.; Huang, B.; Jacob, T. Phys. Chem. Chem. Phys. 2013, 15, 8789–8794. In this ref., it states that the "Only the resonant part of the Bethe–Salpeter Hamiltonian is considered, i.e., Tamm–Dancoff approximation", why I cannot use this approximation for the same 2D silicene system?

[Daniele Varsano]

Dear Zhou Liu-Jiang,
having both em1s and em1d looks to cause problems, we will investigate. Anyway either you have calculated before the em1d, otherwise
just calculate the static one that is the one you need. If you have calculated the pp before, yambo should recognize it and skip the em1s calculations and reading the screening from the database.

why I cannot use this approximation for the same 2D silicene system?

Because you have to remove the fractional occupation first, unfortunately I cannot help on this, did you try to play with the Occupation threshold?

Best,

Daniele

[Davide Sangalli]

(2) If I generate yambo.in by unadding -p p, this job can be done well. While, this "[WARNING]Fractional e/h occupations. Causal BSEmode forced." still existed, what's the reason. I note a reference: Wei, W.; Dai, Y.; Huang, B.; Jacob, T. Phys. Chem. Chem. Phys. 2013, 15, 8789–8794. In this ref., it states that the "Only the resonant part of the Bethe–Salpeter Hamiltonian is considered, i.e., Tamm–Dancoff approximation", why I cannot use this approximation for the same 2D silicene system?

Dear Zhou Liu-Jiang,
the Tamm-Dancoff approximations (TDA) means that "the coupling part of the Hamiltonian is neglected" not that "only the resonant part is considered".
http://www.yambo-code.org/theory/1402-4 ... 09_015.pdf

Only for systems with a gap (i.e. integer occupations) the anti-resonant part make no contribution and the TDA is equal to consider only the resonat part of the BSE Hamiltonian. For metallic systems (i.e. in case of fractional occupations) the anti-resonant part must also be included. Otherwise, instead of the correct dependence on the single particle occupations:
(f_v - f_c ) = f_v * (1 - f_c ) - f_c * (1 - f_v ) ,
you would neglect the anti-resonant term (which is zero for integer occupations. i.e. f_c=0 and f_v=2)
f_c * (1 - f_v )
Here "f_c" are the occupations of the conduction bands while "f_v" are the occupations of the valence bands.

In case you really want to compute only the resonant contribution to the absorption of a metallic system you can open the file "src/bse/K_eh_setup.F" and comment the lines were BSE_mode="causal" is forced. However I guess that, as Daniele suggests, in the paper they had integer occupations...

Davide

[ljzhou86]

[quote="

why I cannot use this approximation for the same 2D silicene system?

Because you have to remove the fractional occupation first, unfortunately I cannot help on this, did you try to play with the Occupation threshold?
[/quote]

How to remove the fractional occupation first? Remove it in the step of pwscf calculation?