EPW Forum

Dear EPW community,

On the basis of the MGB2 case provided by EPW, I calculated MOB2 with the same space group, but the following error occurred

Bloch2wanp: 22 / 22

Writing Hamiltonian, Dynamical matrix and EP vertex in Wann rep to file

===================================================================
Memory usage: VmHWM = 172Mb
VmPeak = 506Mb
===================================================================

Using uniform q-mesh: 20 20 20
Size of q point mesh for interpolation: 8000
Using uniform MP k-mesh: 20 20 20
Size of k point mesh for interpolation: 968
Max number of k points per pool: 70

Fermi energy coarse grid = 19.151446 eV

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Error in routine efermig (1):
internal error, cannot bracket Ef
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

stopping ...

fermi energy calculated by EPW is 19.151446, is larger than the result calculated by pw.x. this is scf.out

[rencong@mu01 epw]$ grep 'Fermi energy' scf.out
the Fermi energy is 19.1246 ev

epw.in show that:

--
&inputepw
prefix = 'MoB2',
amass(1) = 95.94,
amass(2) = 10.811
outdir = './'

ep_coupling = .true.
elph = .true.
kmaps = .false.
epbwrite = .true.
epbread = .false.

epwwrite = .true.
epwread = .false.

etf_mem = 1

nbndsub = 5,

wannierize = .true.
num_iter = 500
dis_froz_max= 8.4
proj(1) = 'B:pz'
proj(2) = 'f=0.5,1.0,0.5:s'
proj(3) = 'f=0.0,0.5,0.5:s'
proj(4) = 'f=0.5,0.5,0.5:s'

iverbosity = 2

eps_acustic = 2.0 ! Lowest boundary for the phonon frequency
ephwrite = .true. ! Writes .ephmat files used when Eliasberg = .true.

fsthick = 0.4 ! eV
degaussw = 0.10 ! eV
nsmear = 1
delta_smear = 0.04 ! eV

degaussq = 0.5 ! meV
nqstep = 500

eliashberg = .true.

laniso = .true.
limag = .true.
lpade = .true.

conv_thr_iaxis = 1.0d-4

wscut = 1.0 ! eV Upper limit over frequency integration/summation in the Elisashberg eq

nstemp = 1 ! Nr. of temps
temps = 1 ! K provide list of temperetures OR (nstemp and temps = tempsmin tempsmax for even space mode)

nsiter = 500

muc = 0.16

dvscf_dir = '../phonon/save'

nk1 = 6
nk2 = 6
nk3 = 6

nq1 = 6
nq2 = 6
nq3 = 6

mp_mesh_k = .true.
nkf1 = 20
nkf2 = 20
nkf3 = 20

nqf1 = 20
nqf2 = 20
nqf3 = 20
/

scf.in show that

&control
calculation='scf',
prefix='MoB2',
pseudo_dir = '../',
outdir='./',
tprnfor = .true.,
tstress = .true.,
etot_conv_thr = 1.0d-5
forc_conv_thr = 1.0d-4
/
&system
ibrav = 0,
nat= 3,
ntyp = 2,
ecutwfc = 40
smearing = 'mp'
occupations = 'smearing'
degauss = 0.02
/
&electrons
diagonalization = 'david'
mixing_mode = 'plain'
mixing_beta = 0.7
conv_thr = 1.0d-9
/
ATOMIC_SPECIES
B 10.811 B.UPF
Mo 95.94 Mo.UPF
CELL_PARAMETERS (angstrom)
3.0304924100 0.0000000000 0.0000000000
-1.5152462050 2.6244834130 0.0000000000
0.0000000000 0.0000000000 3.3493210000
ATOMIC_POSITIONS (crystal)
Mo 0.0000000000 0.0000000000 0.0000000000
B 0.6666666700 0.3333333300 0.5000000000
B 0.3333333300 0.6666666700 0.5000000000
K_POINTS AUTOMATIC
12 12 12 0 0 0

Dear yq_zhao:

fermi energy calculated by EPW is 19.151446, is larger than the result calculated by pw.x. this is scf.out

[rencong@mu01 epw]$ grep 'Fermi energy' scf.out
the Fermi energy is 19.1246 ev

This is not a problem; since your k coarse grid is 6x6x6 and the k grid in your scf calculation is 12x12x12, you can obtain the different Fermi energy.

I think that the issue results from your Wannierization.
Since Mg is [Ne]3s^2 and Mo is [Kr]4d^5 5s, you should change the projection accordingly.
Currently, since your nbndsub (# of Wannier functions) is smaller than the number of electrons in your system, you encounter the issue with calculation of Fermi energy on fine grids.

Sincerely,

H. Lee

Dear H.Lee:

Thank you for your reply. I have been unable to understand how to set nbndsub.Is it the number of valence electrons in the compound? So MgB2 is where Mg atom gives two valence electrons, and a B atom gives 1.5 valence electrons, because of the covalent bond between B. Right?



yuqi_zhao

Dear yuqi_zhao:

(from scf.out for MgB2)

number of electrons = 8.00

For MgB2, the number of valence electrons is 4 and you can find it in your scf.out; if you perform spin-unpolarized calculations, you need to divide the number (8) indicated in "number of electrons" by 2.

nbndsub is the number of Wannier functions and you need to determine it based on the band manifold of interest. To determine the Fermi energy, nbndsub should be equal to or larger than the number of valence electrons.

I would suggest you to have more experience with the construction of Wannier functions before running EPW.

Sincerely,

H. Lee

Dear H.Lee：

Are you saying that if I care about Mo dxy, Mo dxx, Mo s, B pz, Nbnd should be set to 8? So dis_win _min, dis_win_max, dis_froz_min, dis_froz_max, are the parameters set according to these eight orbits.

yq_zhao