[Pw_forum] Cu PP for LDA+U

[Vo, Trinh (388C)]

Dear All,

I am doing a test calculation for YBa2Cu3O6 compound.  I go through all the Pps on PWSCF website.  I noticed a comment in a link "details" for the PP "Cu-pbe-d-rrkjus.UPF."  a comment,
"The original version of this pseudopotential file contained unnormalized

atomic orbitals, which could be a source of trouble for LDA+U calculations.On April 26, 2006 an updated version, containing normalized atomic orbitals,
was uploaded to the site."

Does it mean that any PP that have unnormalized atomic orbitals will give unreliable results in LDA+U calculation.

I used the PP "Cu-pbe-n-van_gipaw.UPF" for one test calculation, and I got a message in the output

"WARNING: atomic wfc #  6 for atom type 3 has zero norm
     WARNING: atomic wfc #  6 for atom type 4 has zero norm
     WARNING: atomic wfc #  6 for atom type 5 has zero norm
     WARNING: atomic wfc #  2 for atom type 6 has zero norm
     WARNING: atomic wfc #  4 for atom type 6 has zero norm"

I guess the message may be the unnormalized atomic wfc.

However, using both Pps above, I still have metallic solution, while it should have been an insulator.  I guess the problem is not just unnormalized atomic wfc but other things (e.g occupation matrix,…), and I have to figure it out.

Thank you,

Trinh

Computation Group

JPL/CalTech


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20110913/c5560970/attachment.html>