about convergence study

Moderators: gmatteo, bruneval, rigna

Post Reply
jiangx3
Posts: 32
Joined: Fri Apr 30, 2010 9:33 pm

about convergence study

Post by jiangx3 » Mon Jun 21, 2010 5:23 pm

Dear abinitors,

I want to ask a question: if I made a convergence study using Gamma point(0.0 0.0 0.0), could I use these input parameters for other k points in the same grid, which means I would not change the ngkpt and shiftk.

Thanks in advance for your replying!


Best regards,
Xueping Jiang

gonze
Posts: 390
Joined: Fri Aug 14, 2009 8:29 pm

Re: about convergence study

Post by gonze » Tue Jun 22, 2010 6:57 pm

Dear Jiangx,

jiangx3 wrote:I want to ask a question: if I made a convergence study using Gamma point(0.0 0.0 0.0), could I use these input parameters for other k points in the same grid, which means I would not change the ngkpt and shiftk.


In general this is true... But for small systems (with a large Brillouin zone), it might
not be true. At least, you will have the right order of magnitude.

X.

jiangx3
Posts: 32
Joined: Fri Apr 30, 2010 9:33 pm

Re: about convergence study

Post by jiangx3 » Wed Jun 23, 2010 5:19 pm

Dear X,

Thanks for your reply. Actually, I just want to study graphene. When I do convergence study of gw parameters, I used Gamma point (0,0,0). But now I want to calculate the band gap of graphene, which means I have to calculate the energy in K point (-1/3 1/3 0.0). Can the input variable value studied from Gamma point use in the K point?

Thanks and Best regards,
Xueping Jiang

gonze
Posts: 390
Joined: Fri Aug 14, 2009 8:29 pm

Re: about convergence study

Post by gonze » Wed Jun 23, 2010 6:42 pm

jiangx3 wrote:Actually, I just want to study graphene. When I do convergence study of gw parameters, I used Gamma point (0,0,0). But now I want to calculate the band gap of graphene, which means I have to calculate the energy in K point (-1/3 1/3 0.0). Can the input variable value studied from Gamma point use in the K point?


Graphene is a very difficult system for GW. It is doubtful that the screening from the LDA wavefunction
at Gamma only will be meaningful. Did you check the literature ? I guess somebody has already done graphene !
X.

jiangx3
Posts: 32
Joined: Fri Apr 30, 2010 9:33 pm

Re: about convergence study

Post by jiangx3 » Thu Jun 24, 2010 3:37 pm

Dear X,

Yes, I have a literature that talks about the gw calculation of graphene. Ab Initio GW Many-body Effects in Graphene, PRL 101, 226405 (2008) or http://prl.aps.org/pdf/PRL/v101/i22/e226405
In the literature, the band gap at K point for graphene is zero.
What I did was that first I used Gamma point to do convergence study, then I used the new input variables to calculate the energy of K point. But the fact is the gap I got is very large, more than 0.7 eV. So I am wondering if the convergent parameters got from Gamma point can not be used for K point.

Best regards,
Xueping Jiang

Post Reply