[Pw_forum] relaxation of doped system
n16031320 at mail.ncku.edu.tw
Tue May 3 06:12:23 CEST 2016
Thanks for your response, I use Ver.5.3.0 of QE and GGA in the form of PBE.
About the type of pseudopotential, I use PAW from THEOS, according to the website it said that PAW type is widely tested and no error is reported.
I study some paper which is experiment, they dope with Ag and the percentage up to 7%, so I think 6.25% may be ok.
I studied the paper which is about experiment, it is said that SnSe doped with Al, Pb and Cu for 2%,respectively, don't change lattice parameter too much,
so I decided to do the relax calculation instead of vc-relax first for doped with Pb and Al for 6.25%, and the result show that the position of atoms don't change too mush,
then I check the total force is <0.0004 , I think the result of relaxation is reasonable.
First I use 12 processors for the vc-relax calculation, now I use 48 processors for the relax calculation,
the relax calculation for the doped with Al and Pb for 6.25% is done, but doped with Cu for 6.25% is still running, maybe 6.25% is too large for Cu?
Also I will try to dope with lower percent too.
And what things I should pay attention to, any advice? Thanks!
Department of Mechanical Engineering
National Cheng Kung University
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the users