Slices of the Fermi surface can, in principle, be calculated in a manor similar to doing band structure calculations by calculating the eigenvalues on a grid setup in the k-points file.  This method is a bit tedious and doesn't always give the nicest looking figures.

The better way would be to use a second program which can interface to VASP or QE like wannier90 (http://www.wannier.org/). Essentially the eigenvalues will be interpolated using Wannier functions to give smooth surfaces which can be viewed in a program like XCrysDen. In their tutorial, they should cover how to do the Fermi surface. Also, wannier90 can be used for much more than that. You will need to recompile VASP with an extra flag to allow for the interface to Wannier90.

In QE you can use  kvecs_FS.x and bands_FS.x. to plot Fermi surface. Check The example02 in QE. there you can find an example for it.

Thanks. let me try wannier90 in VASP code. In QE these commends works. Let me workout and come back if some problem.

Dear All,

How does compile wannier90 within VASP code as Dr. Andrew O'Hara said in his post above.

What is that extra flag ? I have VASP 5.3.3 version and I know simple installation with intel compiler but I do not know which extra flag I have to use in compilation ?

Which file I have to plot from VASP (with wannier90) output for visualizing fermi surface .

Thank You All

Shilendra

Dear Masoud Berahman,

I am new user of QE and learning calculation of fermi surfaces from DFT.  Can you send me link of example 02 of QE which you are talking. I searched on Google, I got tutorial of QE and other example. I got confuse which one is for fermi surface calculation.

Thank You

Shilendra

I compiled VASP with Wannier90, and tried to calculate the Fermi surface of silver. As a result, a surface was displayed in the Xcrysden program, a bit like the one that should be, but very distorted. Does anyone know what parameters to choose in order to improve the calculation result?

Dear Dr. Lev Fomin,

I am facing the same problem. If you resolved the same, can you please help me out.

I would be very thankful to you for the same.

I realized that you do not need to do molecular dynamics. It is necessary to strictly determine the lattice symmetry. Then you need to take a lot of Monkhorst-Paсk points, but so that the gamma point is present (that is, the GAMMA parameter in the INСAR) The problem is that Wannier90 does not want to work with a large number of points. I do not exactly remember the limit, 20 * 20 * 20 or something like that. If you do all this, you get a more or less beautiful Fermi surface.

Dear Sir,

thank you for your positive response. But, i have tried with 5*5*5 as well as 20 *20*20, in each case it is distorted.

Please find the following steps and figure out where i am doing wrong

Firstly, when i used LWANNIER90 = .TRUE. in VASP run with the following wannier90.win file

num_wann =    12  ! set to NBANDS by VASP

use_bloch_phases = T

begin unit_cell_cart

     0.0000000     3.2714012     3.2714012

     3.2714012     0.0000000     3.2714012

     3.2714012     3.2714012     0.0000000

end unit_cell_cart

begin atoms_cart

La       0.0000000     0.0000000     0.0000000

Sb       3.2714012     3.2714012     3.2714012

end atoms_cart

mp_grid =     5    5   5

Then, only the .mmn, .eig, and .win file was generated, no .amn file was there.

I proceeded with the above-generated files and run wannier90.x then wannier.chk file was generated.

Then i rerun wannier90.x with wannier.chk and by adding the following flags in the wannier.win file 

fermi_energy = 6.5120 fermi_surface_plot = true Then, wannier.bxsf was generated, but fermi suface is distorted. Can you please suggest me where i am doing wrong. I would be very thankful to you.

Most likely it is not Wannier, but theVASP parameters. It is necessary that the ions do not move. I'll look at your file later.

ok Sir, thank you!