Dear Iurii, As the resulting U are drastically different (>1eV), what is the criterium that we should use to decide wich is the "good" projector for a given system?
Regards, Antoine Jay LAAS-CNRS Toulouse, France Le Mercredi, Août 26, 2020 11:13 CEST, Timrov Iurii <[email protected]> a écrit: Dear Sergey, As Lorenzo said, yes this is expected. Ab initio U depends on the Hubbard manifold, i.e. which projector functions are used (e.g. atomic, orthogonalized atomic, Wannier functions, you name it). So it is absolutely important to report in papers which projector functions were used for the calculation of U. In particular, this is why one should be very careful when comparing U values from different papers. There are also other factors that influence the value of U (pseudopotentials, oxidation state, functional, one-shot or self-consistent calculation). We have some papers in preparation on this topic, which hopefully would clarify a bit more these points. You can also check this web page: http://theossrv1.epfl.ch/Main/DFTHubbard Greetings, Iurii -- Dr. Iurii TIMROV Postdoctoral ResearcherSTI - IMX - THEOS and NCCR - MARVELSwiss Federal Institute of Technology Lausanne (EPFL)CH-1015 Lausanne, Switzerland +41 21 69 34 881http://people.epfl.ch/265334______________________________________________________________________________ From: users <[email protected]> on behalf of Lorenzo Paulatto <[email protected]> Sent: Wednesday, August 26, 2020 10:59:13 AM To: [email protected] Subject: Re: [QE-users] U calculation, different projection type > Is it expected behavior? Yes, kind regards > Thanks, > Sergey -- Lorenzo Paulatto - Paris _______________________________________________ Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso) users mailing list [email protected] https://lists.quantum-espresso.org/mailman/listinfo/users
_______________________________________________ Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso) users mailing list [email protected] https://lists.quantum-espresso.org/mailman/listinfo/users
