[HSE][GaN] Procedure and obtaining bandstructure

Question on input files/tags, interpreting output, etc.

Please check whether the answer to your question is given in the VASP online manual or has been discussed in this forum previously!

Moderators: Moderator, Global Moderator

Post Reply
Message
Author
karol_kawka
Newbie
Newbie
Posts: 2
Joined: Tue Feb 20, 2024 8:44 pm

[HSE][GaN] Procedure and obtaining bandstructure

#1 Post by karol_kawka » Tue Jun 18, 2024 3:12 pm

Hello,
I'm new in VASP and I'm struggling with HSE calculations for 1x1 GaN unit cell.
Before I explain - I check Si tutorial, but there is something I do wrong.

I will present procedure which I use - just tell me if I think properly.
1. DFT calculations - relaxing the structure
INCAR
System = GaN
ISMEAR = 0
SIGMA = 0.1
IBRION = 1
ISIF = 2
NSW = 300
EDIFFG = -0.001
EDIFF = 1E-06

POSCAR
N Ga N Ga
1.0000000000000000
1.5992848386202618 -2.7700425957176233 0.0000000000000000
1.5992848386202618 2.7700425957176233 0.0000000000000000
0.0000000000000000 0.0000000000000000 5.2096578125058270
N Ga
2 2
Direct
0.3333352518831418 0.6666677663231013 0.3773848816157403
0.6666677663231013 0.3333352518831418 0.8773847937232382
0.6666677663231013 0.3333352518831418 0.5007107021921862
0.3333352518831418 0.6666677663231013 0.0007107900846955

0.00000000E+00 0.00000000E+00 0.00000000E+00
0.00000000E+00 0.00000000E+00 0.00000000E+00
0.00000000E+00 0.00000000E+00 0.00000000E+00
0.00000000E+00 0.00000000E+00 0.00000000E+00

KPOINTS
K-Points
0
Monkhorst Pack
11 11 7
0 0 0

After this step I obtain WAVECAR, CHGCAR and IBZKPT.


Then I go to the second step - HSE Calculation (WAVECAR, CHGCAR were used)
INCAR
HSE06 Calculation
LHFCALC= .TRUE. (Activate HF)
AEXX = 0.25 (25% HF exact exchange, adjusted this value to reproduce experimental band gap)
HFSCREEN= 0.2 (Switch to screened exchange, e.g. HSE06)
ALGO = ALL (Electronic Minimisation Algorithm, ALGO=58)
TIME = 0.4 (Timestep for IALGO5X)
PRECFOCK= N (HF FFT grid)
# NKRED = 2 (Reduce k-grid-even only, see also NKREDX, NKREDY and NKREDZ)
# HFLMAX = 4 (HF cut-off: 4d, 6f)
# LDIAG = .TRUE. (Diagnolise Eigenvalues)

KPOINTS
Automatically generated mesh
72
Reciprocal lattice
0.00000000000000 0.00000000000000 0.00000000000000 1
0.09090909090909 0.00000000000000 0.00000000000000 6
0.18181818181818 0.00000000000000 0.00000000000000 6
0.27272727272727 0.00000000000000 0.00000000000000 6
0.36363636363637 0.00000000000000 0.00000000000000 6
0.45454545454545 0.00000000000000 0.00000000000000 6
0.09090909090909 0.09090909090909 0.00000000000000 6
0.18181818181818 0.09090909090909 0.00000000000000 12
0.27272727272727 0.09090909090909 0.00000000000000 12
0.36363636363637 0.09090909090909 0.00000000000000 12
0.45454545454545 0.09090909090909 0.00000000000000 6
0.18181818181818 0.18181818181818 0.00000000000000 6
0.27272727272727 0.18181818181818 0.00000000000000 12
0.36363636363637 0.18181818181818 0.00000000000000 12
0.27272727272727 0.27272727272727 0.00000000000000 6
0.36363636363637 0.27272727272727 0.00000000000000 6
0.00000000000000 0.00000000000000 0.14285714285715 2
0.09090909090909 0.00000000000000 0.14285714285715 12
0.18181818181818 0.00000000000000 0.14285714285715 12
0.27272727272727 0.00000000000000 0.14285714285715 12
0.36363636363637 0.00000000000000 0.14285714285715 12
0.45454545454545 0.00000000000000 0.14285714285715 12
0.09090909090909 0.09090909090909 0.14285714285715 12
0.18181818181818 0.09090909090909 0.14285714285715 24
0.27272727272727 0.09090909090909 0.14285714285715 24
0.36363636363637 0.09090909090909 0.14285714285715 24
0.45454545454545 0.09090909090909 0.14285714285715 12
0.18181818181818 0.18181818181818 0.14285714285715 12
0.27272727272727 0.18181818181818 0.14285714285715 24
0.36363636363637 0.18181818181818 0.14285714285715 24
0.27272727272727 0.27272727272727 0.14285714285715 12
0.36363636363637 0.27272727272727 0.14285714285715 12
0.00000000000000 0.00000000000000 0.28571428571429 2
0.09090909090909 0.00000000000000 0.28571428571429 12
0.18181818181818 0.00000000000000 0.28571428571429 12
0.27272727272727 0.00000000000000 0.28571428571429 12
0.36363636363637 0.00000000000000 0.28571428571429 12
0.45454545454545 0.00000000000000 0.28571428571429 12
0.09090909090909 0.09090909090909 0.28571428571429 12
0.18181818181818 0.09090909090909 0.28571428571429 24
0.27272727272727 0.09090909090909 0.28571428571429 24
0.36363636363637 0.09090909090909 0.28571428571429 24
0.45454545454545 0.09090909090909 0.28571428571429 12
0.18181818181818 0.18181818181818 0.28571428571429 12
0.27272727272727 0.18181818181818 0.28571428571429 24
0.36363636363637 0.18181818181818 0.28571428571429 24
0.27272727272727 0.27272727272727 0.28571428571429 12
0.36363636363637 0.27272727272727 0.28571428571429 12
0.00000000000000 0.00000000000000 0.42857142857143 2
0.09090909090909 0.00000000000000 0.42857142857143 12
0.18181818181818 0.00000000000000 0.42857142857143 12
0.27272727272727 0.00000000000000 0.42857142857143 12
0.36363636363637 0.00000000000000 0.42857142857143 12
0.45454545454545 0.00000000000000 0.42857142857143 12
0.09090909090909 0.09090909090909 0.42857142857143 12
0.18181818181818 0.09090909090909 0.42857142857143 24
0.27272727272727 0.09090909090909 0.42857142857143 24
0.36363636363637 0.09090909090909 0.42857142857143 24
0.45454545454545 0.09090909090909 0.42857142857143 12
0.18181818181818 0.18181818181818 0.42857142857143 12
0.27272727272727 0.18181818181818 0.42857142857143 24
0.36363636363637 0.18181818181818 0.42857142857143 24
0.27272727272727 0.27272727272727 0.42857142857143 12
0.36363636363637 0.27272727272727 0.42857142857143 12
0.0000000000 0.0000000000 0.0000000000 0.000 GAMMA
0.5000000000 0.0000000000 0.0000000000 0.000 M
0.3333333333 0.3333333333 0.0000000000 0.000 K
0.0000000000 0.0000000000 0.0000000000 0.000 GAMMA
0.0000000000 0.0000000000 0.5000000000 0.000 A
0.5000000000 0.0000000000 0.5000000000 0.000 L
0.3333333333 0.3333333333 0.5000000000 0.000 H
0.0000000000 0.0000000000 0.5000000000 0.000 A

K-ponts where took from IBZKPT file, in the end I added symmetry points, like in tutorial, and I increased the number at the top.

Here are my questions:
1. Is this procedure looks good? I know that paramterization for relaxation is really simple, but I need to understand first some basics. Right now HSE calculation is running on HPC.
2. How can I get bandstructure and DOS after HSE calculations? Do i need ANOTHER calculation with ICHARG = 11, LORBIT = 11? Do I need to use py4vasp or there is another method? As I understand to plot bandstructure I need to have PROCAR, yes?

I have a lot of problems with Si HSE calculation in tutorial, because it produces the same bandstructure, in 'normal' calculation and 'hse'.

Thanks in advance,
Karol.

pedro_melo
Global Moderator
Global Moderator
Posts: 111
Joined: Thu Nov 03, 2022 1:03 pm

Re: [HSE][GaN] Procedure and obtaining bandstructure

#2 Post by pedro_melo » Tue Jun 25, 2024 7:30 pm

Dear karol_kawka,

I do not understand what you are asking here. Is there an issue with your calculations for both GaN and Si? Your input files seem OK, but without concrete information it might be hard to help. Would you mind posting here the input and output files with the issues you are experiencing?

Kind regards,
Pedro

karol_kawka
Newbie
Newbie
Posts: 2
Joined: Tue Feb 20, 2024 8:44 pm

Re: [HSE][GaN] Procedure and obtaining bandstructure

#3 Post by karol_kawka » Fri Jun 28, 2024 1:00 pm

Sorry, I did some more tests with a little success, but questions stays - I will try better... so:

1. In the second step, when I have WAVECAR and and want to start some HSE calculations, I should insert option LORBIT =11 to obtain bandstructure, with mentioned KPOINTS (72 pkt)? So, the HSE procedure in VASP to allows obtain bandstructure, needs only WAVECAR from "normal run", INCAR with HSE settings and apropriate KPOINTS (from IBZKPT file + high symmetry points?


2. I did the HSE calculation and obtain structure(using pyprocar) band_GaN.png. I cut non-zero-weighted kpoints from PROCAR to enlarged right side of the graph and obtain band_right.png. PyPROCAR calculated bandgap which is 3.1 eV (exp 3.4). Is there any method to get more "rounded" graph, not so sharp? I tried something like adding "path" to high symmetry points:
Automatically generated mesh
104
Reciprocal lattice
0.00000000000000 0.00000000000000 0.00000000000000 1
0.09090909090909 0.00000000000000 0.00000000000000 6
0.18181818181818 0.00000000000000 0.00000000000000 6
0.27272727272727 0.00000000000000 0.00000000000000 6
0.36363636363637 0.00000000000000 0.00000000000000 6
0.45454545454545 0.00000000000000 0.00000000000000 6
0.09090909090909 0.09090909090909 0.00000000000000 6
0.18181818181818 0.09090909090909 0.00000000000000 12
0.27272727272727 0.09090909090909 0.00000000000000 12
0.36363636363637 0.09090909090909 0.00000000000000 12
0.45454545454545 0.09090909090909 0.00000000000000 6
0.18181818181818 0.18181818181818 0.00000000000000 6
0.27272727272727 0.18181818181818 0.00000000000000 12
0.36363636363637 0.18181818181818 0.00000000000000 12
0.27272727272727 0.27272727272727 0.00000000000000 6
0.36363636363637 0.27272727272727 0.00000000000000 6
0.00000000000000 0.00000000000000 0.14285714285715 2
0.09090909090909 0.00000000000000 0.14285714285715 12
0.18181818181818 0.00000000000000 0.14285714285715 12
0.27272727272727 0.00000000000000 0.14285714285715 12
0.36363636363637 0.00000000000000 0.14285714285715 12
0.45454545454545 0.00000000000000 0.14285714285715 12
0.09090909090909 0.09090909090909 0.14285714285715 12
0.18181818181818 0.09090909090909 0.14285714285715 24
0.27272727272727 0.09090909090909 0.14285714285715 24
0.36363636363637 0.09090909090909 0.14285714285715 24
0.45454545454545 0.09090909090909 0.14285714285715 12
0.18181818181818 0.18181818181818 0.14285714285715 12
0.27272727272727 0.18181818181818 0.14285714285715 24
0.36363636363637 0.18181818181818 0.14285714285715 24
0.27272727272727 0.27272727272727 0.14285714285715 12
0.36363636363637 0.27272727272727 0.14285714285715 12
0.00000000000000 0.00000000000000 0.28571428571429 2
0.09090909090909 0.00000000000000 0.28571428571429 12
0.18181818181818 0.00000000000000 0.28571428571429 12
0.27272727272727 0.00000000000000 0.28571428571429 12
0.36363636363637 0.00000000000000 0.28571428571429 12
0.45454545454545 0.00000000000000 0.28571428571429 12
0.09090909090909 0.09090909090909 0.28571428571429 12
0.18181818181818 0.09090909090909 0.28571428571429 24
0.27272727272727 0.09090909090909 0.28571428571429 24
0.36363636363637 0.09090909090909 0.28571428571429 24
0.45454545454545 0.09090909090909 0.28571428571429 12
0.18181818181818 0.18181818181818 0.28571428571429 12
0.27272727272727 0.18181818181818 0.28571428571429 24
0.36363636363637 0.18181818181818 0.28571428571429 24
0.27272727272727 0.27272727272727 0.28571428571429 12
0.36363636363637 0.27272727272727 0.28571428571429 12
0.00000000000000 0.00000000000000 0.42857142857143 2
0.09090909090909 0.00000000000000 0.42857142857143 12
0.18181818181818 0.00000000000000 0.42857142857143 12
0.27272727272727 0.00000000000000 0.42857142857143 12
0.36363636363637 0.00000000000000 0.42857142857143 12
0.45454545454545 0.00000000000000 0.42857142857143 12
0.09090909090909 0.09090909090909 0.42857142857143 12
0.18181818181818 0.09090909090909 0.42857142857143 24
0.27272727272727 0.09090909090909 0.42857142857143 24
0.36363636363637 0.09090909090909 0.42857142857143 24
0.45454545454545 0.09090909090909 0.42857142857143 12
0.18181818181818 0.18181818181818 0.42857142857143 12
0.27272727272727 0.18181818181818 0.42857142857143 24
0.36363636363637 0.18181818181818 0.42857142857143 24
0.27272727272727 0.27272727272727 0.42857142857143 12
0.36363636363637 0.27272727272727 0.42857142857143 12
0.333333 0.333333 0.000000 0.000 K
0.315789 0.315789 0.000000 0.000
0.298246 0.298246 0.000000 0.000
0.280702 0.280702 0.000000 0.000
0.263158 0.263158 0.000000 0.000
0.245614 0.245614 0.000000 0.000
0.228070 0.228070 0.000000 0.000
0.210526 0.210526 0.000000 0.000
0.192982 0.192982 0.000000 0.000
0.175439 0.175439 0.000000 0.000
0.157895 0.157895 0.000000 0.000
0.140351 0.140351 0.000000 0.000
0.122807 0.122807 0.000000 0.000
0.105263 0.105263 0.000000 0.000
0.087719 0.087719 0.000000 0.000
0.070175 0.070175 0.000000 0.000
0.052632 0.052632 0.000000 0.000
0.035088 0.035088 0.000000 0.000
0.017544 0.017544 0.000000 0.000
0.000000 0.000000 0.000000 0.000 GAMMA
0.000000 0.000000 0.000000 0.000 GAMMA
0.000000 0.000000 0.026316 0.000
0.000000 0.000000 0.052632 0.000
0.000000 0.000000 0.078947 0.000
0.000000 0.000000 0.105263 0.000
0.000000 0.000000 0.131579 0.000
0.000000 0.000000 0.157895 0.000
0.000000 0.000000 0.184211 0.000
0.000000 0.000000 0.210526 0.000
0.000000 0.000000 0.236842 0.000
0.000000 0.000000 0.263158 0.000
0.000000 0.000000 0.289474 0.000
0.000000 0.000000 0.315789 0.000
0.000000 0.000000 0.342105 0.000
0.000000 0.000000 0.368421 0.000
0.000000 0.000000 0.394737 0.000
0.000000 0.000000 0.421053 0.000
0.000000 0.000000 0.447368 0.000
0.000000 0.000000 0.473684 0.000
0.000000 0.000000 0.500000 0.000 A

but it turns out with graphs like morekp2.png.
PyPROCAR calculate bandgap as a 0eV, so there is some error, for sure.

3. To conclude, is there any procedure in VASP to obtain DOS and Bandstructure at the same time? I use PyPTOCAR and vaspkit for post-process... in Vaspkit I can plot DOS, but in PyPROCAR there is an error "'NoneType' object has no attribute 'energies'". I know that is not your thing, so if I couldn't fix it, I will use Vaspkit for DOS, and PyPROCAR for bandstructure. I had a problems with compile VASP on PC with HDF library, tried lots of times, I abaddoned it.

My first post was about to get sure that my procedure was correct :)
Karol.
You do not have the required permissions to view the files attached to this post.

Post Reply