[Pw_forum] Will the variance of pseudopotential lead to quite different results in the terms of total energy ?

[Nicola Marzari]

Dear Yuanqing,


they seem to have used the GBRV set - this is available here
https://www.physics.rutgers.edu/gbrv/, so the wisest thing to
do would be use those, first, and make sure you get the same results.

If you don't, maybe ask the authors to send you the pseudopotential files.
If you still do not get the same results, maybe you are doing something 
different
with cutoffs, sampling, smearing, etc...

Your calculations are done at U=0; note that for U different from zero
your pseudopotential matters a lot, since the atomic manifold to which you
apply the U term can be quite different (especially if the pseudos where 
generated
in different states of charge) - so different calculations become even 
less comprable,
because you cannot use the same U.

			nicola
On 02/11/2015 14:06, ＷＡＮＧ ＹＵＡＮＱＩＮＧ wrote:
> Dear all,
>
>    Hi!!!
>    I want to reproduce the results of one article (J. Phys. Chem. C, 2015, 119 (9), pp 4827–4833) because the authors kindly provide the raw input. Since they use the pseudopotential I can not find in the official website, I just change them to the pseudopotential in the official website. However, I found the results are quite different, even in the sense of relative energy. So, I wonder how I should choose the correct or reasonable psudopotential.
>
>            total_energy(my results)           total_energy(JPCC)
> case 1    -251.52904868 Ry(0.)            -500.88392739 Ry(0.)
> case 2    -251.86326485 Ry(-0.33421617)   -501.15752582 Ry(-0.27359843)
> case 3    -252.00820845 Ry(-0.47915977)   -501.26364613 Ry(-0.37971874)
> case 4    -252.04128086 Ry(-0.51223218)   -501.27151855 Ry(-0.38759116)
> case 5    -252.00878190 Ry(-0.47973322)   -501.22233815 Ry(-0.33841076)
>
>
> ******input file***************
> &CONTROL
>   calculation = 'vc-relax'
>   pseudo_dir = '/espresso-5.1.2/pseudo/'
> /
> &SYSTEM
>   ibrav = 0
>   nat = 6
>   ntyp = 2
>   nbnd = 78
>   starting_magnetization(1) = 0.6
>   starting_magnetization(2) = 0
>   ecutwfc = 40
>   ecutrho = 500
>   occupations = 'smearing'
>   degauss = 0.01
>   smearing = 'mp'
>   nspin = 2
>   Hubbard_U(1) = 0
>   Hubbard_U(2) = 0
>   Hubbard_alpha(1) = 0
>   Hubbard_alpha(2) = 0
> /
> &ELECTRONS
> /
> &IONS
> /
> &CELL
>   cell_dofree = 'shape'
> /
> K_POINTS {automatic}
>   5 5 5 0 0 0
> ATOMIC_SPECIES
>   Ir0 192.22 Ir.pbe-n-rrkjus.UPF    #ir_pbe_v1.uspp.F.UPF in the article
>   O1 15.9994 O.pbe-n-rrkjus_psl.0.1.UPF  #o_pbe_v1.01.uspp.F.UPF
> ATOMIC_POSITIONS {crystal}
>   Ir0 0.00000 0.00000 0.00000 0 0 0
>   Ir0 0.50000 0.50000 0.50000 0 0 0
>   O1 0.30500 0.30500 0.00000 1 1 0
>   O1 -0.30500 -0.30500 0.00000 1 1 0
>   O1 0.80500 0.19500 0.50000 1 1 0
>   O1 0.19500 0.80500 0.50000 1 1 0
> CELL_PARAMETERS {angstrom}
>   4.25368570224 0.0 0.0
>   0.0 4.25368570224 0.0
>   0.0 0.0 2.74126411922
> ****************************************
>
> Thank you very much!!!
>
> Best regards,
>
> Yuanqing Wang
>
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://pwscf.org/mailman/listinfo/pw_forum
>

-- 
----------------------------------------------------------------------
Prof Nicola Marzari, Chair of Theory and Simulation of Materials, EPFL
Director, National Centre for Competence in Research NCCR MARVEL, EPFL
http://theossrv1.epfl.ch/Main/Contact http://nccr-marvel.ch/en/project