HSE npt using GW
Posted: Tue May 05, 2015 7:38 pm
Dear Abinit developers,
I have a question regarding the HSE exchange correlation potential.
I've looked at the earlier version of the Abinit exchange input files (t41.in and t42.in) and it seemed like the only way to use HSE was by using gwcalctype using the GW methodolgy, however later on in later versions one was able to use external libxc
library by specifying the ixc= (-524 or -428) for HSE and HSE06 without going to GW. My question is how to use properly HSE without getting into GW ?
Why i am asking this is because i've tested perovskite MAPbCl3 bandgap and seemed like HSE did not do anything to correct the gap, i.e. stayed as SOC bandgap without the change.
I've also tried to use SOC with GW and it failed saying i cannot do so now (usewvl variable) which i understand is not implemented yet.
P.S. Calculation speed seems to be same as compared to no-exchange at all.
I have a question regarding the HSE exchange correlation potential.
I've looked at the earlier version of the Abinit exchange input files (t41.in and t42.in) and it seemed like the only way to use HSE was by using gwcalctype using the GW methodolgy, however later on in later versions one was able to use external libxc
library by specifying the ixc= (-524 or -428) for HSE and HSE06 without going to GW. My question is how to use properly HSE without getting into GW ?
Why i am asking this is because i've tested perovskite MAPbCl3 bandgap and seemed like HSE did not do anything to correct the gap, i.e. stayed as SOC bandgap without the change.
I've also tried to use SOC with GW and it failed saying i cannot do so now (usewvl variable) which i understand is not implemented yet.
P.S. Calculation speed seems to be same as compared to no-exchange at all.