[QE-users] Questions on obtaining Hubbard U and J by linear response

[ZHANG Yichi]

Dear all,

I want to use the methods in PHYSICAL REVIEW B 71, 035105 (2005) and PHYSICAL REVIEW B 84, 115108 (2011) to obtain the U and J in my system. My system is a CsPbCl3 supercell where one singlet Pb(II) atom is replaced by a sextet Mn(II) atom whose five d electrons are parallel. I need to find the U and J of the Mn atom.

According to the tutorial (https://hjkgrp.mit.edu/tutorials/2011-05-31-calculating-hubbard-u/), I need to know the occupation numbers of the Mn d orbitals. The result of projwfc.x shows two d shells of the Mn atom, one halfly occupied (spin-up: 0.9, 0.9, 0.9, 0.9, 0.9; spin-down 0.1, 0.1, 0.1, 0.1, 0.1) and the other one unoccupied (both spin-up and spin-down: 0.05, 0.05, 0.05, 0.05 ,0.05, the numbers are only for demonstration). My first question: Should I consider both d shells or only the halfly occupied one for each alpha/beta perturbation? Or does this mean that EACH d shell should have its OWN U and J?

My second question is on how to obtain the occupation numbers of the first SCF iteration for bare response. As far as I know, pw.x does not export the wavefunctions to pwscf.save until the job is done, so I wonder how to have the unconverged wavefunctions in the pwscf.save.

I would appreciate it if anyone helps me with these two questions.

Yichi Zhang
Postgraduate student
Department of Chemistry
The Hong Kong University of Science and Technology
Office 4207, Clearwater Bay, New Territories, Hong Kong
+852 5219 8714

[cid:6a0499e6-f99a-4d33-9f64-57963789ed75]
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20240805/52d7e174/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Outlook-1u1twpgz.jpg
Type: image/jpeg
Size: 86684 bytes
Desc: Outlook-1u1twpgz.jpg
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20240805/52d7e174/attachment.jpg>