force_symmorphic Breaking Symmetry

[chowdhury.cua]

Dear All,

I am a new user of Yambo. Yambo is really good code and I had practiced all the tutorial and it worked very well.

For my work I am using Yambo. My system is SnS. I have used force_symmorphic=.true. for scf and nscf calculations and in output I have noticed that symmetry is breaking and it is going to C1 symmetry. I have tried at 20 times to restore the symmetry and I got the same results. Below I have mentioned my input please find it.

Please let me know you suggestions.

Thank you for your time and help.

Best,
Sugata

Input :- Scf:-
CONTROL
calculation="scf",
restart_mode='from_scratch',
outdir = '/tmp/SnS_VC_Bulk_SRL/'
prefix = 'SnS_VC_Bulk_SRL',
pseudo_dir ='/home/snc8/2D-Materials/Sn/S/Band/Relax/GWA/NoShift',
forc_conv_thr = 1.0D-4,
etot_conv_thr = 1.0D-3,
wf_collect=.true.
verbosity = 'high'
nstep = 5000

/
&SYSTEM
ibrav = 8,
celldm(1) = 21.164931072
celldm(2) = 0.35625,
celldm(3) = 0.3875,
nat = 2,
ntyp = 2,
!nbnd = 120,
lspinorb = .true.,
noncolin = .true.,
ecutwfc = 50.D0,
ecutrho = 300.D0,
force_symmorphic=.true.
occupations='smearing',
smearing='marzari-vanderbilt',
degauss=0.02

/
&ELECTRONS
diago_full_acc=.true.,
diagonalization='david'
mixing_mode = 'local-TF'
mixing_beta = 0.5,
conv_thr = 1.0D-8

/
&IONS
upscale = 100

/
&CELL
cell_factor = 1.8D0

/
!CELL_PARAMETERS {alat}
! 1.00000000 0.00000 0.0000
! 0.00000000 0.35625 0.0000
! 0.00000000 0.00000 0.3875
ATOMIC_SPECIES
Sn 118.71 sn_frl_gga.upf
S 32.065 s_frl_gga.upf

K_POINTS {automatic}
12 12 12 0 0 0

ATOMIC_POSITIONS {crystal}
.......

[myrta gruning]

Dear Sugata,

First note that the inputs you posted are with pwscf. pwscf is a different code, it is interfaced with yambo, but for any issue with that code, you should refer to the pwscf mailing list.
If the system you are studying has nonsymmorphic symmetries and you force to use only symmorphic ones, the code will recognize the system has less symmetries as it should. In principle this should not affect the results, though there may be small differences due to the numerical error.

Is the difference only in the number of symmetries or did you obtain results for the electronic structure (KS energies, total energies) that are significantly different when using or not the nonsymmorphic ops in the DFT part?
If it is with the DFT part you should refer to the pwscf mailing list (see above) for help/advice.
If it is with yambo what kind of calculations are you running and where did they differ from what you would expect? In that case please upload log/report/outputs


Best
m

[chowdhury.cua]

Dear Myrta,

Thank you for your reply. Actually I didn't think about the nonsymmorphic symmetries. SnS is a nonsymmorphic system. I will work on it.

Also, in my my previous post question was not write. My question is due to symmetry breaking the GE bandstructure will be significant different than DFT bandstructure.

Thank you,
Sugata

[myrta gruning]

Dear Sugata

that I know there is nothing special from the GW point of view for nonsymmorphic system. With respect to DFT, usually the GW opens a gap and changes slightly the band-width.
The reason you have to force the system to be symmorphic (e.g. discarding nonsymmorphic symmetries) is just technical: nonsymmorphic symmetry ops are not implemented in yambo.

Best