[QE-developers] Pseudo-core density with NCPPs

[Jacob Williams]

Dear all,

Is it possible to obtain the (pseudo-orbital approximation to the) all-electron density from a norm-conserving pseudopotential calculation? I thought at first that rho_core ought to provide the rest of the approximated density, but (rho%of_r + rho_core) doesn’t seem to integrate to the total number of electrons in the small case I tried, Si with 2x2x2 k-points: rho%of_r integrates to 4, but rho%of_r + rho_core to somewhat less than 4.75.

Is rho_core only for nonlinear core correction? Must one use projector-augmented waves or ultrasoft PPs to get explicit core-electron properties?

Of course, I do not expect the core density to be very accurate! But I do need to access some effect from the core anyway. If it helps, I’m using norm-conserving Vanderbilt optimized PPs from abinit and PBE, with scalar relativistic correction; my ecutwfc = 75 Ry. I’m also doing LSDA calculations, so the actual quantity I’ve been looking at is 0.5 * (rho_core(:) + rho%of_r(:, 1) +/- rho%of_r(:, 2)).

Thanks for your time!

Sincerely yours,

Jacob Williams
Ph.D. candidate, Yang group
Duke University Department of Chemistry
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/developers/attachments/20230602/bb86f029/attachment.html>