I am a beginner with quantum espresso, and am trying to relax a sulfite ion. For some reason I cannot get it to converge. Here is my input file:
&CONTROL
calculation='relax'
prefix='sulfite'
pseudo_dir = '/home/asmaldo1/SSSP_efficiency_pseudos',
outdir = './outdir'
/
&SYSTEM
ibrav=1,
celldm(1)=20,
nat=4,
ntyp=2,
ecutwfc=75
tot_charge=-2
/
&ELECTRONS
/
&IONS
/
&CELL
cell_dofree='ibrav'
/
ATOMIC_SPECIES
S 32.065 s_pbe_v1.4.uspp.F.UPF
O 15.9994 O.pbe-n-kjpaw_psl.0.1.UPF
ATOMIC_POSITIONS (angstrom)
S -0.0001 -0.0345 -0.5997
O -1.2433 0.7050 0.1933
O 1.2423 0.7069 0.1932
O 0.0011 -1.3774 0.2131
K_POINTS (automatic)
1 1 1 0 0 0
Is there something that I am missing, or something else that I could try? Thanks.
Hello my friend,
you don't need to use the block &CELL, because you are running in 'relax' mode.
Without the output is a little bit hard to understand what is happening, but anyway, i'd try to use this flags:
&ELECTRONS
mixing_beta = 0.25 ,
diagonalization = 'cg' ,
the first flag will reduce your density matrix mixing factor for self-consistency from 0.7 (qe default) to 0.25 and the second flag changes the diagonalization to the conjugate-gradient-like band-by-band diagonalization (is very slower than, but lightly more robust than the default 'david').
try decrease the ecutwfc too
Hope this helps
Best regards
@ Gustavo Silvestre Thank you, my sulfite ion converged successfully. I previously had attempted to change the mixing beta to 0.3, yet it still did not converge. The key was changing the diagonalization. I see your answer some what comes from the documentation of QE, yet is there a way you could explain why is works? If not, thank you for making it work out!
You have choose very less number of k-points as 1 1 1.
First of all Increase this value and also increase convergence thread.
I will suggest you to check with the examples of relaxation under pWSCF examples
- Badges
- Science topic