<div dir="ltr"><div><div>Dear Stefan nqx in your first calc is set to 4 while it is set to 1 in the second. Only one q for forck is most probably too scarce. It might be the reason for the discrepancy.<br><br></div>cheers<br><br></div>Layla<br></div><div class="gmail_extra"><br><div class="gmail_quote">2018-05-06 17:34 GMT+02:00 Dr. Thomas Brumme <span dir="ltr"><<a href="mailto:thomas.brumme@uni-leipzig.de" target="_blank">thomas.brumme@uni-leipzig.de</a>></span>:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Dear Stefan,<br>
<br>
you can't calculate the band structure using a self-consistent calculation.<br>
This would mean that you search for the ground-state density (scf) whilst<br>
your integrations in k-space are done on a very sparse grid... On the other<br>
hand you can't calculate the band structure for hybrid functionals using<br>
an nscf or bands calculation - the answer was given recently by Lorenzo<br>
Paulatto and I cite his answer at the end of the mail - you can find it<br>
under the following link:<br>
<br>
<a href="https://www.mail-archive.com/users@lists.quantum-espresso.org/msg33817.html" rel="noreferrer" target="_blank">https://www.mail-archive.com/u<wbr>sers@lists.quantum-espresso.or<wbr>g/msg33817.html</a><br>
<br>
I can furthermore add to Lorenzos answer that you can also get the band<br>
structure by interpolating the E(k) dispersion on the regular grid like it<br>
is done in the BoltzTraP code:<br>
<br>
<a href="https://www.imc.tuwien.ac.at//forschungsbereich_theoretische_chemie/forschungsgruppen/prof_dr_gkh_madsen_theoretical_materials_chemistry/boltztrap/" rel="noreferrer" target="_blank">https://www.imc.tuwien.ac.at//<wbr>forschungsbereich_theoretische<wbr>_chemie/forschungsgruppen/<wbr>prof_dr_gkh_madsen_<wbr>theoretical_materials_chemistr<wbr>y/boltztrap/</a><br>
<a href="https://www.sciencedirect.com/science/article/pii/S0010465506001305" rel="noreferrer" target="_blank">https://www.sciencedirect.com/<wbr>science/article/pii/S001046550<wbr>6001305</a><br>
<br>
Regards<br>
<br>
Thomas<br>
<br>
<br>
the mail:<br>
<br>
Hello,<br>
The problem is quite fundamental, because in order to get the Fock operator at a certain k-point you need the wavefunctions on a grid that is commensurate with it, this can only be done self-consistently.<br>
<br>
However, there are a few things you can do.<br>
<br>
1. I think you can add k-points to the scf calculation with weight zero, this is boring (you'll have to specify he grid and the path by hand) and very computationally expensive.<br>
<br>
2. you can interpolate the Fock operator, this should be possible if it is expressed om a basis of localized wavrfunctions. There is some development in this direction, but I don't know if it has been implemented already by someone.<br>
<br>
3. You could interpolate the wavefunctions, from Wannier functions. This is probably impossible in practice.<br>
<br>
4. You can interpolate the Hamiltonian, passing through a Wannier basis representation. This is possible, I have done it, and I've put an example in the git developement branch (PP/examples/W90_open_grid_exa<wbr>mple/). But you need to know, or learn, how to use Wannier90 (which is a useful skill regardless), and to use the developement version of qe.<br>
<br>
Zitat von Stefan Seidl <<a href="mailto:Stefan.Seidl@physik.uni-regensburg.de" target="_blank">Stefan.Seidl@physik.uni-regen<wbr>sburg.de</a>>:<div><div class="h5"><br>
<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
Dear All,<br>
I am calculating with QUANTUM ESPRESSO the BANDSTRUCTURE of Silicon with HYBRID FUNCTIONAL HSE.<br>
<br>
My calculation contains 3 steps:<br>
<br>
1: pw.x < <a href="http://si.scf.in" rel="noreferrer" target="_blank">si.scf.in</a> > si.scf.out<br>
2: pw.x < <a href="http://si.scf_instead_bands.in" rel="noreferrer" target="_blank">si.scf_instead_bands.in</a> > si.scf_instead_bands.out<br>
3: bands.x < <a href="http://bands.in" rel="noreferrer" target="_blank">bands.in</a> > bands.out<br>
<br>
to 1: I am using an explicit kpoint list with a correspoinding qpoint grid<br>
all kpoints (12 12 12 mesh )are equal weighted with 1 (which provides the weight 2/(12**3) )<br>
##############################<wbr>#########################<br>
&control<br>
calculation = 'scf'<br>
prefix = 'silicon',<br>
pseudo_dir = './',<br>
outdir = './calculation'<br>
verbosity = 'high',<br>
/<br>
&system<br>
ibrav = 2,<br>
celldm(1) = 10.2612,<br>
nat = 2,<br>
ntyp = 1,<br>
ecutwfc = 40.0,<br>
! nbnd = 8,<br>
input_dft ='hse',<br>
nqx1 = 4,<br>
nqx2 = 4,<br>
nqx3 = 4,<br>
occupations = 'smearing',<br>
smearing = 'gaussian',<br>
degauss = 0.001,<br>
<br>
nosym = .TRUE.,<br>
noinv = .TRUE.,<br>
/<br>
&electrons<br>
mixing_beta = 0.7<br>
/<br>
<br>
ATOMIC_SPECIES<br>
Si 28.0855 Si_NCv0.4_PBE_stringent.upf<br>
ATOMIC_POSITIONS<br>
Si 0.00 0.00 0.00<br>
Si 0.25 0.25 0.25<br>
K_POINTS tpiba<br>
1728<br>
0.0000000 0.0000000 0.0000000 1<br>
-0.0833333 0.0833333 -0.0833333 1<br>
-0.1666667 0.1666667 -0.1666667 1<br>
-0.2500000 0.2500000 -0.2500000 1<br>
-0.3333333 0.3333333 -0.3333333 1<br>
-0.4166667 0.4166667 -0.4166667 1<br>
0.5000000 -0.5000000 0.5000000 1<br>
0.4166667 -0.4166667 0.4166667 1<br>
0.3333333 -0.3333333 0.3333333 1<br>
0.2500000 -0.2500000 0.2500000 1<br>
0.1666667 -0.1666667 0.1666667 1<br>
0.0833333 -0.0833333 0.0833333 1<br>
0.0833333 0.0833333 0.0833333 1<br>
0.0000000 0.1666667 0.0000000 1<br>
-0.0833333 0.2500000 -0.0833333 1<br>
-0.1666667 0.3333333 -0.1666667 1<br>
...<br>
<br>
<br>
<br>
<br>
to 2: here I am just specifying the special kpoint path<br>
##############################<wbr>#########################<br>
<br>
&control<br>
calculation = 'scf'<br>
prefix = 'silicon',<br>
pseudo_dir = './',<br>
outdir = './calculation'<br>
verbosity = 'high',<br>
/<br>
&system<br>
ibrav = 2,<br>
celldm(1) = 10.2612,<br>
nat = 2,<br>
ntyp = 1,<br>
ecutwfc = 40.0,<br>
nbnd = 8,<br>
input_dft ='hse',<br>
nqx1 = 1,<br>
nqx2 = 1,<br>
nqx3 = 1,<br>
occupations = 'smearing',<br>
smearing = 'gaussian',<br>
degauss = 0.001,<br>
<br>
nosym = .TRUE.,<br>
noinv = .TRUE.,<br>
/<br>
&electrons<br>
mixing_beta = 0.7<br>
/<br>
<br>
ATOMIC_SPECIES<br>
Si 28.0855 Si_NCv0.4_PBE_stringent.upf<br>
ATOMIC_POSITIONS<br>
Si 0.00 0.00 0.00<br>
Si 0.25 0.25 0.25<br>
K_POINTS tpiba<br>
50<br>
0.50000000 0.50000000 0.50000000 1<br>
0.44444444 0.44444444 0.44444444 2<br>
0.38888889 0.38888889 0.38888889 3<br>
0.33333333 0.33333333 0.33333333 4<br>
0.27777778 0.27777778 0.27777778 5<br>
0.22222222 0.22222222 0.22222222 6<br>
0.16666667 0.16666667 0.16666667 7<br>
0.11111111 0.11111111 0.11111111 8<br>
0.05555556 0.05555556 0.05555556 9<br>
0.00000000 0.00000000 0.00000000 10<br>
0.00000000 0.00000000 0.00000000 11<br>
...<br>
<br>
<br>
<br>
The eigenvalues from my 1st calculation si.scf.out are correct !<br>
But the eigenvalues of the 2nd calculation si.bands.out are wrong !<br>
E.g. the HOMO-LUMO distance at the Gamma point is just too big (compared to VASP results which are proven to be right)!<br>
<br>
Where is the mistake in my input file <a href="http://si.bands.in" rel="noreferrer" target="_blank">si.bands.in</a> for the 2nd calculation ??<br>
<br>
Thanks a lot, best<br>
Stefan<br>
</blockquote>
<br></div></div>
-- <br>
Dr. rer. nat. Thomas Brumme<br>
Wilhelm-Ostwald-Institute for Physical and Theoretical Chemistry<br>
Leipzig University<br>
Phillipp-Rosenthal-Strasse 31<br>
04103 Leipzig<br>
<br>
Tel: +49 (0)341 97 36456<br>
<br>
email: <a href="mailto:thomas.brumme@uni-leipzig.de" target="_blank">thomas.brumme@uni-leipzig.de</a><br>
<br>
______________________________<wbr>_________________<br>
users mailing list<br>
<a href="mailto:users@lists.quantum-espresso.org" target="_blank">users@lists.quantum-espresso.o<wbr>rg</a><br>
<a href="https://lists.quantum-espresso.org/mailman/listinfo/users" rel="noreferrer" target="_blank">https://lists.quantum-espresso<wbr>.org/mailman/listinfo/users</a><br>
</blockquote></div><br></div>