[gpaw-users] Advice sought to change coordinates and functional in dielectric response code
Ask Hjorth Larsen
asklarsen at gmail.com
Wed Aug 29 16:17:48 CEST 2018
Dear Subhabrata Das,
Den ons. 25. jul. 2018 kl. 22.38 skrev Subhabrata Das via gpaw-users
<gpaw-users at listserv.fysik.dtu.dk>:
>
> Hi GPAW Users
>
> I am new to GPAW and am trying to use the following script which is for measuring dielectric function of Graphene:
>
>
> a = 2.5
>
> c = 3.22
>
>
> pbc = [True, True, False]
>
>
> GR = Atoms(symbols='C2', positions=[(0.5 * a, -np.sqrt(3) / 6 * a, 0.0),
>
> (0.5 * a, np.sqrt(3) / 6 * a, 0.0)],
>
> cell=[(0.5 * a, -0.5 * 3**0.5 * a, 0),
>
> (0.5 * a, 0.5 * 3**0.5 * a, 0),
>
> (0.0, 0.0, c * 2.0)], pbc=True)
>
> old_cell = GR.get_cell()
>
> old_cell[2, 2] = 2 * c
>
> GR.set_cell(old_cell)
>
> atoms = GR
>
>
> calc = GPAW(h=0.18,
>
> mode=PW(400),
>
> kpts={'density': 10.0, 'gamma': True},
>
> occupations=FermiDirac(0.1))
>
>
> atoms.set_calculator(calc)
>
> atoms.get_potential_energy()
>
> calc.write('gs.gpw')
>
>
> kpts = find_high_symmetry_monkhorst_pack('gs.gpw', density=30,
>
> pbc=pbc)
>
> responseGS = GPAW('gs.gpw',
>
> fixdensity=True,
>
> kpts=kpts,
>
> parallel={'band': 1},
>
> nbands=30,
>
> occupations=FermiDirac(0.001),
>
> convergence={'bands': 20})
>
>
> responseGS.get_potential_energy()
>
> responseGS.write('gsresponse.gpw', 'all')
>
>
> df = DielectricFunction('gsresponse.gpw', eta=25e-3,
>
> pbc=pbc, domega0=0.01,
>
> integrationmode='tetrahedron integration')
>
> df1tetra, df2tetra = df.get_dielectric_function(q_c=[0, 0, 0])
>
>
> df = DielectricFunction('gsresponse.gpw',
>
> domega0=0.01,
>
> eta=25e-3)
>
> df1, df2 = df.get_dielectric_function(q_c=[0, 0, 0])
>
> omega_w = df.get_frequencies()
>
>
>
>
>
> I intend to use the compute the dielectric function for another structure with different coordinates.
>
> define the coordinates the atoms
>
> %
>
> xyzmat = [
>
> -6.31297563 0.26518140 -0.00051415
>
> -3.87247004 0.04624524 -0.00011671
>
> -2.54471461 -0.45092975 0.00025424
>
> -1.40762214 0.32164685 0.00025716
>
> -0.14082181 -0.36737620 0.00040491
>
> -1.41631871 1.83241036 -0.00028412
>
> 1.08500702 0.22931606 0.00011588
>
> 2.31251187 -0.52667176 0.00029553
>
> 3.58450542 -0.03439531 -0.00002190
>
> 4.69242411 -0.99804716 0.00016087
>
> 3.93186597 1.43544603 -0.00063508
>
> 6.00605553 -0.69633629 -0.00043540
>
> -4.91747033 -0.71481386 -0.00006503
>
> -6.33877541 0.81302732 -0.00150725
>
> -6.80250908 0.64585156 -0.88455325
>
> -6.80259297 -0.64424678 0.88417077
>
> -4.77255469 -1.70750561 0.00029893
>
> -2.43481082 -1.52152059 0.00053665
>
> -0.18223036 -1.44294275 0.00056995
>
> -0.89955135 2.20480824 -0.87659139
>
> -0.89895740 2.20542498 0.87539859
>
> -2.40695624 2.26194989 -0.00006578
>
> 1.14705911 1.29977283 -0.00045117
>
> 2.19590646 -1.59737031 0.00067991
>
> 4.40950930 -2.03715494 0.00077876
>
> 4.52349738 1.68398373 0.87376327
>
> 3.06269923 2.07763491 -0.00034371
>
> 6.74491659 -1.47529960 -0.00026410
>
> 4.52256678 1.68347836 -0.87581179
>
> 6.74491659 -1.47529960 -0.00026410
>
> 6.37440888 0.31214595 -0.00106985
>
> ];
>
>
> with the supercell size being :
>
> [30 0 0; 0 15 0; 0 0 6];
>
>
> How do I plug in the coordinates of the new structure to replace the coordinates in the gpaw code to compute the dielectric response of the new material? Also how do I change the exchange correlation functional in the code (LDA by default) to another functional? I would greatly appreciate your help. Thanks.
>
Please see the description of the GPAW keywords in the manual, e.g.
the xc keyword.
As for the coordinates, you need to get them as an ASE atoms object,
probably using ase.io.read(filename).
Best regards
Ask
>
> Sincerely
>
> Subhabrata
>
>
>
>
> --
> Subhabrata Das
> PhD Candidate
> Langmuir Center of Colloids and Interfaces
> Columbia University in the City of New York
>
>
>
>
>
> _______________________________________________
> gpaw-users mailing list
> gpaw-users at listserv.fysik.dtu.dk
> https://listserv.fysik.dtu.dk/mailman/listinfo/gpaw-users
More information about the gpaw-users
mailing list