<div><div dir="auto">Dear Prof Peter Blaha,</div></div><div dir="auto"><br></div><div dir="auto">Thank you very much for your detailed explanation.</div><div dir="auto"><br></div><div dir="auto">Many thanks,</div><div dir="auto">Ding</div><div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Mon, 27 Apr 2020 at 4:25 pm, Peter Blaha <<a href="mailto:pblaha@theochem.tuwien.ac.at">pblaha@theochem.tuwien.ac.at</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">For a difference density it is of course necessary that you treat the <br>
same states in the solids and in the atom.<br>
While in an atom this is fairly simple, because each state is classified <br>
by n,l,m (or kappa), for the solid we are missing the "n". We know <br>
easily the "l" (s,p,d,f), but not the principal quantum number n (like <br>
4p or 5p).<br>
<br>
This can be only be done by comparising with the atom, via the <br>
"knowledge" what a chemist would call "valence and core" states and via <br>
the energies of these states.<br>
<br>
a) execute: x lstart -sigma<br>
b) This creates a case.inst_sigma input file, where the "P" indicates <br>
to you, which atomic densities are plotted.<br>
c) Look into case.outputst to see their energies (and in particular <br>
those of lower lying states (semi-core states) which have a "T" <br>
indicating that they are included in the solid as "valence" and some of <br>
them must be cut away.<br>
d) now look at your scf1 file (both the E-parameters as you showed us, <br>
but also the eigenvalues. Sometimes also the band-ranges in case.output2 <br>
are helpful.:<br>
<br>
> ATOMIC SPHERE DEPENDENT PARAMETERS FOR ATOM Ce<br>
> :e__0001: OVERALL ENERGY PARAMETER IS 0.5943<br>
> OVERALL BASIS SET ON ATOM IS LAPW<br>
> :E2_0001: E( 2)= 0.5943<br>
> APW+lo<br>
> :E2_0001: E( 2)= -6.9944 E(BOTTOM)= -7.002 E(TOP)= -6.987 <br>
1 2 170<br>
> LOCAL ORBITAL<br>
> :E0_0001: E( 0)= 0.5943<br>
> APW+lo<br>
> :E0_0001: E( 0)= -1.8665 E(BOTTOM)= -2.558 E(TOP)= -1.175 <br>
4 5 176<br>
> LOCAL ORBITAL<br>
> :E1_0001: E( 1)= 0.9943<br>
> APW+lo<br>
> :E1_0001: E( 1)= -0.4363 E(BOTTOM)= -1.448 E(TOP)= 0.575 <br>
3 4 197<br>
> LOCAL ORBITAL<br>
> :E3_0001: E( 3)= 0.6959 E(BOTTOM)= 0.466 E(TOP)= 0.925 <br>
0 1 123<br>
> APW+lo<br>
<br>
There should be Ce-d states around -7 Ry. Compare to the outpust file <br>
and you should see that these are 4d states (they are a bit lower in the <br>
atom) and thus "not valence", so they sould be cut away.<br>
You can do the same for Ce-s (5s) and finally 5p. The latter are at -1.6 <br>
in the atom, but according to your list at -0.4 in the solid.<br>
<br>
e) So look again at the eigenvalues of case.output1 or the band ranges <br>
in output2 and you should find a solution: most likely your emin <br>
should be around -0.2 or so, i.e. in the "gap" between the real valence <br>
states (B-sp Ce-4f,5d,6sp) and the semicore states as identified above.<br>
<br>
PS: In some cases, the "automatic" procedure with "x lstart -sigma" may <br>
not work. Take a Ce-oxide, there will be O-2s states at LOWER energy <br>
than these Ce-5p states and you CANNOT select an EMIN, which "keeps" the <br>
O-2s but cuts away the Ce-5p. In such cases you have to explicitly <br>
include / exlude certain states in the atom (eg. edit case.inst_sigma <br>
and replace the "P" by "N" for the O-2s states, rerun:<br>
lstart lstart.def (where lstart.def should contain: case.inst_sigma, <br>
not just case.inst).<br>
<br>
Am 27.04.2020 um 04:38 schrieb Ding Peng:<br>
> Dear WIEN2k experts,<br>
> <br>
> I am trying to calculate the difference electron density of CeB6, which has a space group of Pm-3m, by WIEN2k (16.04). After I finished running SCF cycles (I used non-magnetic calculation GGA+U calculation, runsp_c_lapw, by forcing the magnetic moment in a spin-polarised setup to zero), I followed the electron density calculation process:<br>
> <br>
> 1) check case.scf1, it shows<br>
> <br>
> ---------------------------------------------------------------------<br>
> LDA+U potential added for atom type 1 L= 3 spin up<br>
> :LMAX-WF: 10 Non-Spherical LMAX: 4<br>
> <br>
> ATOMIC SPHERE DEPENDENT PARAMETERS FOR ATOM Ce<br>
> :e__0001: OVERALL ENERGY PARAMETER IS 0.5943<br>
> OVERALL BASIS SET ON ATOM IS LAPW<br>
> :E2_0001: E( 2)= 0.5943<br>
> APW+lo<br>
> :E2_0001: E( 2)= -6.9944 E(BOTTOM)= -7.002 E(TOP)= -6.987 1 2 170<br>
> LOCAL ORBITAL<br>
> :E0_0001: E( 0)= 0.5943<br>
> APW+lo<br>
> :E0_0001: E( 0)= -1.8665 E(BOTTOM)= -2.558 E(TOP)= -1.175 4 5 176<br>
> LOCAL ORBITAL<br>
> :E1_0001: E( 1)= 0.9943<br>
> APW+lo<br>
> :E1_0001: E( 1)= -0.4363 E(BOTTOM)= -1.448 E(TOP)= 0.575 3 4 197<br>
> LOCAL ORBITAL<br>
> :E3_0001: E( 3)= 0.6959 E(BOTTOM)= 0.466 E(TOP)= 0.925 0 1 123<br>
> APW+lo<br>
> <br>
> ATOMIC SPHERE DEPENDENT PARAMETERS FOR ATOM B<br>
> :e__0002: OVERALL ENERGY PARAMETER IS 0.5943<br>
> OVERALL BASIS SET ON ATOM IS LAPW<br>
> :E0_0002: E( 0)= 0.5943<br>
> APW+lo<br>
> :E1_0002: E( 1)= 0.5943<br>
> APW+lo<br>
> ---------------------------------------------------------------------<br>
> <br>
> <br>
> 2) I choose the default value Emin = -1 and rerun LAPW2 calculation<br>
> <br>
> 3) run lstart -sigma to calculate the superposed atomic electron density<br>
> <br>
> 4) run lapw5 and set the option "DIFF" in case.in5<br>
> <br>
> <br>
> <br>
> <br>
> The result difference electron density map was presented by Xcrysden. I found very strong electron accumulation (delta rho is positive) near the core of Ce ions and electron depletion(delta rho is negative) near the core of B ions.<br>
> <br>
> My questions are:<br>
> <br>
> 1) Does this result makes sense? Because Ce3+ is cation, which should tend to lose electrons, and therefore delta rho should be negative near the core of Ce. This contradicts to my difference electron density map.<br>
> <br>
> 2) Is the superposed atomic electron density equivalent to the superposed atomic electron densities from independent atomic model (IAM)? Or it is equivalent to the superposition of the atomic electron densities of ions?<br>
> <br>
> 3) Is it correct to set Emin = -1? I had read the past discussions about this topic from the Wien2k email list, bu<a href="https://www.google.com/maps/search/t+still+have+no+idea+how+to+determine+E?entry=gmail&source=g">t still have no idea how to determine E</a>min based on the results in case.scf1.<br>
> <br>
> <br>
> Look forward to the answers.<br>
> <br>
> Many thanks,<br>
> Ding<br>
> <br>
> _______________________________________________<br>
> Wien mailing list<br>
> <a href="mailto:Wien@zeus.theochem.tuwien.ac.at" target="_blank">Wien@zeus.theochem.tuwien.ac.at</a><br>
> <a href="http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien" rel="noreferrer" target="_blank">http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien</a><br>
> SEARCH the MAILING-LIST at: <a href="http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html" rel="noreferrer" target="_blank">http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html</a><br>
> <br>
<br>
-- <br>
Peter Blaha<br>
Inst.Materials Chemistry<br>
TU Vienna<br>
Getreidemarkt 9<br>
A-1060 Vienna<br>
Austria<br>
+43-1-5880115671<br>
_______________________________________________<br>
Wien mailing list<br>
<a href="mailto:Wien@zeus.theochem.tuwien.ac.at" target="_blank">Wien@zeus.theochem.tuwien.ac.at</a><br>
<a href="http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien" rel="noreferrer" target="_blank">http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien</a><br>
SEARCH the MAILING-LIST at: <a href="http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html" rel="noreferrer" target="_blank">http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html</a><br>
</blockquote></div></div>