[Pw_forum] Convergence of RS-GaAs

[Gabriele Sclauzero]

naphtaly moro wrote:
> Dear All,
> I am trying to do an SCF calculation on GaAs with a Rocksalt
> structure. The calculations does not converge, I have tried to
> increase the Ecut up to 200Ryd but still there's no convergence. I
> have also attempted to increase the k-points with no success. What
> might be the problem?

I don't know if LDA or GGA is able to correctly predict the insulating GaAs ground state. 
Maybe the LDA gap is very small and you need to use broadening of the occupations to 
converge the calculations.
Try to use occupations='smearing' in your input file.


>  My input is
> 
>  &control
>     calculation = 'scf'
>     restart_mode='from_scratch',
>     prefix='GaAs',
>     pseudo_dir='./'
>     outdir = './',
>  /
>  &system
>     ibrav=  2, celldm(1) =10.7, nat=  2, ntyp= 2,
>     ecutwfc = 24, ecutrho = 100, 

You must NOT specify ecutrho when using norm-conserving pseudos. ecutrho must be 4* 
ecutwfc (which is the default). It is of no use specifying larger values.

> nbnd = 8, nosym=.true.,

Why do you want to avoid the use of symmetries? It usually helps the convergence.

>  /
>  &electrons
>     diagonalization='david'
>     mixing_mode = 'plain'
>     mixing_beta = 0.3
>     conv_thr =  1.0d-8

>     electron_maxstep=1000

That's not wise to do, if it doesn't reach convergence in 100 steps it is quite likely 
that it will never do...


>  /
> ATOMIC_SPECIES
>  Ga  69.723  ga_ppot.gon
>  As  74.9216 as_ppot.gon

Are you sure that your pseudopotentials are good? This might be another source of 
trouble... have you tried with different pseudos, for instance those on the QE website?


HTH

GS

> ATOMIC_POSITIONS
>  Ga 0.00 0.00 0.00
>  As 0.50 0.50 0.50
> K_POINTS {automatic}
>  4 4 4 0 0 0
> 
> PS: I am using normconserving pseudos
> 

-- 


o ------------------------------------------------ o
| Gabriele Sclauzero, PhD Student                  |
| c/o:   SISSA & CNR-INFM Democritos,              |
|        via Beirut 2-4, 34014 Trieste (Italy)     |
| email: sclauzer at sissa.it                         |
| phone: +39 040 3787 511                          |
| skype: gurlonotturno                             |
o ------------------------------------------------ o