[Pw_forum] CuO, LDA+U, Pseudopots, etc.

[Paul M. Grant]

To All (especially Paolo, Nicola, Matteo and Axel):

I’m in the middle of attempting, as an exercise, the calculation of the
electronic structure of monoclinic CuO (tenorite) within the LDA+U
provisions of PWscf (Cu is on the “approved list” of elements contained in
set_hubbard_l.f90 and tabd.f90).  FeO and NiO I get to go OK, but CuO is
giving me problems.  Tenorite is a well-established AF Hubbard-type
insulator, and the LDA+U approach can yield a supportive band structure
(e.g., see JPCM 11, 5021 (1999) and references therein).  However, I am not
aware of an LDA+U calculation starting from a PW/PP formalism for tenorite.
If one exists, please send me the reference and ignore the rest of this
posting.

Simplistically, the initial d-state configuration of Cu2+ in CuO by Hund’s
Rule should be 3d9 (one empty minority spin state state
a hole).   But, with
nspin=2, lda_plus_u set true, and U=5 eV (arbitrary), the initial minority
spin occupation gets set to unity, not 4/5 on average (3/5 for Ni and 1/5
for Fe).  Is this due to occ_loc being set to 10 in tabd.f90 and not 9 (it’s
6 for Fe and 8 for Ni)?  The occupation distribution converges to around
0.92 for eg and 0.99 for t2g with self-consistency, a metallic state, and no
amount of “alternative user ns eigenvalue guidance” gives a gap.  I haven’t
tried any Wannier projections
rich man’s or poor
yet, because I suspect
there is another issue
choice of the proper Cu PP.

The choice of Cu pseudopotentials available in the PWscf library are rather
limited and weird
at least to my naïve and limited experience.  I have been
using primarily the Perdew-Zunger  LDA PP, Cu.pz-n-van_ak.UPF, computed with
valence orbital occupations 3d9.5, 4s1.0 and 4p0.5 (!) (the PW91 and PBE
functionals have equivalent occupations and give similar results).  When I
try Cu.pz-d-rrkjus.UPF, with 3d10, 4s1 and 4p-1(!!!), the first scf
iteration yields individual spin occupations of nearly 5(!!!), and the
situation gets really ugly thereafter.

Why aren’t there Cu PPs with occupations 3d10, 4s1, or even 3d9, 4s2?  I’ve
spent the afternoon studying Paolo’s guide to PP generation and would be
willing to give these a try if that would help
but I sorely need some
advice.  I’ve also been looking over David Vanderbilt’s web site and found
this bit of wisdom:  “Choosing the method of pseudizing the local
potential
This is something of a black art”.

-Paul

PS
there is something truly marvelous and mysterious about the CuO double
bond
it’s given us high temperature superconductivity.  But 21 years later,
there’s no general agreement on wherein and why it arises.

PPS
Is there a utility somewhere that plots the radial dependence of V(r)
and/or rho(r) of a given PP?

Paul M. Grant, PhD
Principal, W2AGZ Technologies
Visiting Scholar, Applied Physics, Stanford University
EPRI Science Fellow (Retired)
IBM Research Staff Member Emeritus
w2agz at pacbell.net
http://www.w2agz.com
 
 


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20071212/066b3dab/attachment.html>