<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">
<META NAME="Generator" CONTENT="MS Exchange Server version 6.5.7654.12">
<TITLE>Work function and void level using exchange-correlation potential</TITLE>
</HEAD>
<BODY>
<!-- Converted from text/plain format -->

<P><FONT SIZE=2>Dear PW_forum users,<BR>
<BR>
I was working with the post-processing of qe to obtain the work functions for some metallic slabs (aluminum, gold..); as also explained in one of the examples of PP, I've plotted the potential of the slab as a function of the out-of-plane axis to obtain the potential in the void region.<BR>
<BR>
My problem rises up when I compare the total potential (local + Hartree + XC as in plot_num=1 ) with the electrostatic one (local + Hartree as in plot_num=11 ). The two potentials are quite different from each other even in the void region, where the XC contribute should(?) be zero.<BR>
(here a plot I have obtained for an Aluminum slab    <A HREF="https://dl.dropboxusercontent.com/u/38867209/Aluminum_6layer111slab_LDA.ps">https://dl.dropboxusercontent.com/u/38867209/Aluminum_6layer111slab_LDA.ps</A> )<BR>
<BR>
I understand that the XC contributions are more long-ranged than the electrostatic ones and this explains the different shape of the two curves, what I would like to understand is why is there this constant shift when the potentials reach the plateau (and, more important, which one is better to trust).  The shift is certainly non-negligible in the systems that I've analyzed (0.41eV in Al_LDA, 0.44eV in Al_GGA, 0.19eV in Au_LDA and 1.12eV in Au_GGA) and the total potential shows a more "noisy" behaviour with an odd curvature (the system are all neutral).<BR>
<BR>
I tried several cell size to see if the problem is related to the amount of void but it seems unrelated (in the plot that I've linked to you the slab distance is set to 30 A, that I think is quite enough to avoid any periodic effect). Same thing for the k_points sampling<BR>
<BR>
I'm working with version 5.0.2 of qe (I've also tried with 4.3.1 obtaining same results)<BR>
<BR>
Since the problem seems to be related to the exchange part I also tried other kinds of pseudos.. the above data comes from US PP, so i tried a norm conserving one for Al_LDA and the shift got reduced by an order of magnitude (the strange bent behaviour still remained)<BR>
<BR>
My question is..  has anyone ever dealt with this kind of issue? it's a known problem or it has just popped up in my calculations? <BR>
<BR>
I hope I've made myself clear and I thank you in advance for your consideration. If there is something else that you need I will try to produce it to you as soon as possible.<BR>
Best regards,<BR>
<BR>
Marco<BR>
<BR>
---<BR>
Marco Pividori<BR>
PhD Student<BR>
Department of Physics<BR>
University of Trieste<BR>
Strada Costiera 11<BR>
I-34151 Trieste<BR>
<BR>
<BR>
<BR>
<BR>
</FONT>
</P>

</BODY>
</HTML>