[gpaw-users] FW: FW: GLLB-SC in LCAO mode

Wed Apr 13 20:59:36 CEST 2016

Dear Xiaoming

As Korina says, the derivative discontinuity should be zero for a metal. Recall that it is just a single number and cannot in general be used to contruct quasiparticle band structures - allthough it may sometimes be a good approximation to use it as a rigid shift of the conduction bands.

Thus, if you have an insulator in the vicinity of a metal you cannot use it to correct the insulator bands. Even if you had access to the exact derivative discontinuity it should be zero in that case, which just tells you that the exact quasiparticle gap of the combined system vanishes. Note, however, that the GLLBSC Kohn-Sham gap is also somewhat larger than what is found with LDA (see the paper you mention). For the case of h-BN/graphene, this may capture part of the proximity effect, and seems to be in agreement with GW calculations, but it is difficult to say how accurate that approach is in general

/Thomas
________________________________________
Fra: gpaw-users-bounces at listserv.fysik.dtu.dk [gpaw-users-bounces at listserv.fysik.dtu.dk] p&#229; vegne af Korina Kuhar via gpaw-users [gpaw-users at listserv.fysik.dtu.dk]
Sendt: 13. april 2016 17:27
Til: Xiaoming Wang; gpaw-users at listserv.fysik.dtu.dk
Emne: Re: [gpaw-users] FW:  FW:  GLLB-SC in LCAO mode

Hi again,

No, since Dxc is a potential.

>From what l can see in the paper ( didn't have the chance to read it through) they are applying the Dxc in a similar way: they shift valnce bands down for 1/2Dxc and cb up for 1/2Dxc therefore opening the gap. Hm, metals? Not sure but don't think it should affect metals, it cannot open a gap if initially Kohn Sham gap doesn't exist. But so far l only used it for semiconductors so I'm not the expert.
Hope it helps.

\K.

Korina Kuhar
PhD student
DTU Physics

Technical University of Denmark

Department of Physics
Fysikvej
Building 307
2800 Kgs. Lyngby
Denmark
Direct +45 45253293
kokuha at fysik.dtu.dk
www.fysik.dtu.dk

________________________________________
From: Xiaoming Wang [xw213 at rci.rutgers.edu]
Sent: Wednesday, April 13, 2016 3:02 PM
To: Korina Kuhar; gpaw-users at listserv.fysik.dtu.dk
Subject: RE: [gpaw-users] FW:  FW:  GLLB-SC in LCAO mode

Thanks Korina. So the Dxc term cannot be applied to the Hamiltonian, right?
Then, how about the band structure of metals or metal-semiconductor hybrids?
I did see a paper PHYSICAL REVIEW B 86, 045208 (2012), which calculate the
band structure of graphene-hBN hybrid system. So in this case, the Dxc only
apply for the hBN bands?

Best,
Xiaoming
Rutgers U

-----Original Message-----
From: Korina Kuhar [mailto:Kokuha at fysik.dtu.dk]
Sent: Wednesday, April 13, 2016 7:17 AM
To: Xiaoming Wang; gpaw-users at listserv.fysik.dtu.dk
Subject: RE: [gpaw-users] FW: FW: GLLB-SC in LCAO mode

Hi!

When calling the get_eigenvalues function the values you get are without
Dxc, so when plotting your band structure you should manually add Dxc to the
unoccupied bands.

Best,
Korina.

Korina Kuhar
PhD student
DTU Physics

Technical University of Denmark

Department of Physics
Fysikvej
Building 307
2800 Kgs. Lyngby
Denmark
Direct +45 45253293
kokuha at fysik.dtu.dk
www.fysik.dtu.dk

________________________________________
From: gpaw-users-bounces at listserv.fysik.dtu.dk
[gpaw-users-bounces at listserv.fysik.dtu.dk] on behalf of Xiaoming Wang via
gpaw-users [gpaw-users at listserv.fysik.dtu.dk]
Sent: Wednesday, April 13, 2016 1:11 AM
To: gpaw-users at listserv.fysik.dtu.dk
Subject: [gpaw-users] FW:  FW:  GLLB-SC in LCAO mode

Hi,

Luckily, I see the examples of diamond in the test directory. For my case, I
try to calculate the band structure of WSe2 by GLLBSC. Following is my
script. However, the obtained band structure is the one without adding the
discontinuity potential. What's the problem? Does anyone have experiences
with bands calculation by GLLBSC functional?
----------------------------------------------------------------------------
-----
from gpaw import *
from ase import *
from ase.io import read
from ase.dft.kpoints import ibz_points, get_bandpath import numpy as np

atoms = read('wse2.cif')
atoms.set_pbc([1,1,1])

calc = GPAW(mode='lcao',
            nbands=18,
            kpts={'size':(12,12,1),'gamma':True},
            xc='GLLBSC',
            h=0.18,
            occupations=FermiDirac(width=0.01),
            convergence={'density':1.0e-4},
            maxiter=200,
            txt='gs_scf.txt',
            setups={'W':'6'},
            basis='dzp',
            verbose=0
            )

atoms.set_calculator(calc)
atoms.get_potential_energy()
calc.write('wse2_gs.gpw')
homolumo = calc.occupations.get_homo_lumo(calc.wfs)
response = calc.hamiltonian.xc.xcs['RESPONSE']
response.calculate_delta_xc(homolumo=homolumo)
#EKs, Dxc = response.calculate_delta_xc_perturbation()
#Gap = EKs + Dxc
calc.write('wse2_gllbsc.gpw')

nbands=16
calc = GPAW('wse2_gllbsc.gpw',
            nbands=18,
            fixdensity=True,
            symmetry='off',
            txt='nscf_bands.txt',
            convergence={'bands': nbands}) G = [0., 0., 0.] K = [1/3., 1/3.,
0.] M = [0.5, 0., 0.] kpts, x, X = get_bandpath([G, M, K, G],
calc.atoms.cell, npoints=100)
calc.set(kpts=kpts)
calc.get_potential_energy()
ef = calc.get_fermi_level()
e_kn = np.array([calc.get_eigenvalues(k) for k in range(len(kpts))]) e_kn -=
ef f = open('bands.dat','w') for n in range(nbands):
    for k in range(len(kpts)):
       print >> f, x[k], e_kn[k,n]
    print >> f

-------------------------------------------------------------------------

Best,
Xiaoming
Postdoc
Rutgers U

-----Original Message-----
From: gpaw-users-bounces at listserv.fysik.dtu.dk
[mailto:gpaw-users-bounces at listserv.fysik.dtu.dk] On Behalf Of Xiaoming Wang
via gpaw-users
Sent: Tuesday, April 12, 2016 4:24 PM
To: gpaw-users at listserv.fysik.dtu.dk
Subject: Re: [gpaw-users] FW: GLLB-SC in LCAO mode

Hi,

The gpaw-1.0.0 works for GLLBSC in LCAO mode now, I used the wrong version.
So is it possible to get the band structure with GLLBSC, not only the band
gap?

Best,
Xiaoming

-----Original Message-----
From: gpaw-users-bounces at listserv.fysik.dtu.dk
[mailto:gpaw-users-bounces at listserv.fysik.dtu.dk] On Behalf Of Xiaoming Wang
via gpaw-users
Sent: Tuesday, April 12, 2016 4:10 PM
To: gpaw-users at listserv.fysik.dtu.dk
Subject: [gpaw-users] FW: GLLB-SC in LCAO mode

Hi,

With gpaw-1.0.0, I got the same error with LCAO mode. Is it fixed?

By the way, is it possible to get the band structure with GLLBSC, not only
the band gap?

Best,
Xiaoming Wang
Postdoc
Rutgers U

-----Original Message-----
From: gpaw-users-bounces at listserv.fysik.dtu.dk
[mailto:gpaw-users-bounces at listserv.fysik.dtu.dk] On Behalf Of Ask Hjorth
Larsen
Sent: Wednesday, September 2, 2015 12:18 PM
To: Christopher Patrick
Cc: gpaw-users at listserv.fysik.dtu.dk
Subject: Re: [gpaw-users] GLLB-SC in LCAO mode

Fixed in r13143.

Please verify it carefully, comparing to a similar real-space calculation.

Best regards
Ask

2015-09-02 17:12 GMT+02:00 Christopher Patrick <chripa at fysik.dtu.dk>:
> Dear all,
>
> I would be interested to know whether one can perform calculations
> with the GLLB-SC functional in LCAO mode.
>
> Based on the attached simple script and the latest GPAW version from
> the svn, it would seem that one cannot; the calculation fails when
> calculating the derivative discontinuity with the error,
>
> File "si.py", line 19, in <module>
>     EKs, Dxc = response.calculate_delta_xc_perturbation()
>   File
"/home/niflheim/chripa/gpaw_latest/gpaw/gpaw/xc/gllb/c_response.py", line
367, in calculate_delta_xc_perturbation
>     gaps.append(self.calculate_delta_xc_perturbation_spin(s))
>   File
"/home/niflheim/chripa/gpaw_latest/gpaw/gpaw/xc/gllb/c_response.py", line
328, in calculate_delta_xc_perturbation_spin
>     self.wfs.add_orbital_density(nt_G, kpt, lumo_n)
> AttributeError: LCAOWaveFunctions instance has no attribute
'add_orbital_density'
>
> which I guess means that the LCAOWaveFunctions instance has no
> attribute 'add_orbital_density' :(
>
> Thanks for any help,
>
> Christopher Patrick
>
> Postdoc, DTU Physics
> _______________________________________________
> gpaw-users mailing list
> gpaw-users at listserv.fysik.dtu.dk
> https://listserv.fysik.dtu.dk/mailman/listinfo/gpaw-users
_______________________________________________
gpaw-users mailing list
gpaw-users at listserv.fysik.dtu.dk
https://listserv.fysik.dtu.dk/mailman/listinfo/gpaw-users

_______________________________________________
gpaw-users mailing list
gpaw-users at listserv.fysik.dtu.dk
https://listserv.fysik.dtu.dk/mailman/listinfo/gpaw-users

_______________________________________________
gpaw-users mailing list
gpaw-users at listserv.fysik.dtu.dk
https://listserv.fysik.dtu.dk/mailman/listinfo/gpaw-users

_______________________________________________
gpaw-users mailing list
gpaw-users at listserv.fysik.dtu.dk
https://listserv.fysik.dtu.dk/mailman/listinfo/gpaw-users

_______________________________________________
gpaw-users mailing list
gpaw-users at listserv.fysik.dtu.dk
https://listserv.fysik.dtu.dk/mailman/listinfo/gpaw-users