## Problem in Anisotropic superconducting gap

General discussion around the EPW software

Moderator: hlee

ilias-serifi
Posts: 59
Joined: Mon Dec 17, 2018 12:43 pm
Affiliation:

### Problem in Anisotropic superconducting gap

Hi dear users and experts
well i had a several problems when i calculate the anisotropic superconducting gap in the first calculation i had 9 K° in the critical temperature and now is totally different also the calculation need some memory and i have 32go which is fair enough for it
can someone help me to find the good results
best
here attached the last part of the output
===================================================================
Eliashberg Spectral Function in the Migdal Approximation
===================================================================

lambda : 1.2556402
lambda_tr : 13.1165215

Estimated Allen-Dynes Tc

logavg = 0.0006359 l_a2f = 1.2921504
mu = 0.10 Tc = 9.808471650344 K
mu = 0.12 Tc = 9.129227645383 K
mu = 0.14 Tc = 8.454883352153 K
mu = 0.16 Tc = 7.787349799824 K
mu = 0.18 Tc = 7.128747303870 K
mu = 0.20 Tc = 6.481416687512 K
a2F : 0.24s CPU 0.24s WALL ( 1 calls)

===================================================================
Solve anisotropic Eliashberg equations
===================================================================

Fermi level (eV) = -3.5501000000E+00
DOS(states/spin/eV/Unit Cell) = 1.4109068289E+00
Electron smearing (eV) = 5.0000000000E-01
Fermi window (eV) = 1.0000000000E+10
Nr k-points within the Fermi shell = 900 out of 900

12 bands within the Fermi window

Max nr of q-points = 225

Electron-phonon coupling strength = 6.4866515

Estimated Allen-Dynes Tc = 4.057478 K for muc = 0.00000

Estimated BCS superconducting gap = 0.615378 meV

WARNING WARNING WARNING

The code may crash since tempsmin = 300.000 K is larger than Allen-Dynes Tc = 4.057 K

temp( 1) = 300.00000 K

Solve anisotropic Eliashberg equations on imaginary-axis

Total number of frequency points nsiw( 1) = 5
Cutoff frequency wscut = 0.8934

Size of allocated memory per pool: ~= 1.8510 Gb
iter ethr znormi deltai [meV]
1 2.279336E+00 6.148708E+00 5.215160E-01
2 1.339930E-01 6.148729E+00 4.775872E-01
3 3.092487E-02 6.148721E+00 4.700531E-01
4 3.973105E-01 6.148721E+00 3.816094E-01
5 8.256705E-02 6.148759E+00 4.011637E-01
6 7.259228E-03 6.148760E+00 3.992406E-01
7 8.316652E-02 6.148771E+00 3.773703E-01
8 3.576812E-02 6.148775E+00 3.677818E-01
9 1.301273E-02 6.148775E+00 3.639366E-01
10 7.597991E-03 6.148777E+00 3.605009E-01
11 8.908209E-03 6.148779E+00 3.566948E-01
12 1.594154E-02 6.148781E+00 3.516557E-01
13 2.915997E-02 6.148784E+00 3.449029E-01
14 2.321005E-02 6.148786E+00 3.401298E-01
15 1.141880E-02 6.148787E+00 3.371709E-01
16 2.830961E-02 6.148790E+00 3.300729E-01
17 1.500637E-02 6.148791E+00 3.254408E-01
18 1.843971E-02 6.148793E+00 3.215237E-01
19 2.570052E-02 6.148795E+00 3.156685E-01
20 8.478597E-03 6.148795E+00 3.149421E-01
21 4.826951E-03 6.148796E+00 3.131499E-01
22 1.981656E-02 6.148798E+00 3.084268E-01
23 3.956379E-02 6.148801E+00 3.003407E-01
24 6.392236E-03 6.148801E+00 2.984998E-01
25 2.638171E-02 6.148803E+00 2.926978E-01
26 2.599393E-02 6.148805E+00 2.867723E-01
27 9.901237E-03 6.148806E+00 2.844691E-01
28 1.510105E-02 6.148807E+00 2.813973E-01
29 9.204302E-03 6.148804E+00 2.819870E-01
30 2.949812E-02 6.148809E+00 2.754994E-01
31 1.216517E-02 6.148809E+00 2.726969E-01
32 2.799378E-02 6.148812E+00 2.675462E-01
33 2.466141E-02 6.148814E+00 2.623058E-01
34 2.113142E-02 6.148815E+00 2.581886E-01
35 1.583407E-02 6.148816E+00 2.549341E-01
36 5.539559E-03 6.148816E+00 2.533955E-01
37 2.246924E-02 6.148818E+00 2.490714E-01
38 1.881116E-02 6.148819E+00 2.452244E-01
39 2.259623E-02 6.148820E+00 2.408130E-01
40 1.292819E-02 6.148821E+00 2.382974E-01
41 1.600064E-02 6.148822E+00 2.352622E-01
42 2.272010E-02 6.148823E+00 2.310276E-01
43 2.714419E-02 6.148824E+00 2.260608E-01
44 3.215957E-03 6.148824E+00 2.254820E-01
45 2.561634E-02 6.148825E+00 2.208533E-01
46 1.520502E-03 6.148825E+00 2.210087E-01
47 2.851552E-02 6.148826E+00 2.159828E-01
48 3.137343E-02 6.148828E+00 2.104584E-01
49 1.815435E-02 6.148829E+00 2.073619E-01
50 1.036164E-02 6.148830E+00 2.056239E-01
51 1.839332E-02 6.148830E+00 2.024942E-01
52 1.414453E-02 6.148830E+00 2.002513E-01
53 1.041734E-02 6.148827E+00 1.985882E-01
54 1.343623E-02 6.148825E+00 1.964410E-01
55 2.288499E-02 6.148833E+00 1.926973E-01
56 2.257672E-02 6.148833E+00 1.891405E-01
57 6.719260E-03 6.148833E+00 1.879992E-01
58 3.573062E-02 6.148833E+00 1.822496E-01
59 3.703434E-03 6.148835E+00 1.818878E-01
60 2.047280E-02 6.148836E+00 1.788538E-01
61 2.709115E-02 6.148837E+00 1.750454E-01
62 2.268132E-02 6.148837E+00 1.718628E-01
63 2.077246E-02 6.148838E+00 1.689050E-01
64 1.209122E-02 6.148838E+00 1.671263E-01
65 1.588000E-02 6.148839E+00 1.648511E-01
66 9.436507E-03 6.148839E+00 1.635016E-01
67 9.937231E-03 6.148839E+00 1.620607E-01
68 1.132592E-02 6.148839E+00 1.604860E-01
69 1.246841E-02 6.148839E+00 1.587051E-01
70 2.153758E-02 6.148839E+00 1.560064E-01
71 1.617998E-02 6.148838E+00 1.539079E-01
72 2.001948E-02 6.148840E+00 1.513090E-01
73 1.679955E-02 6.148842E+00 1.490849E-01
74 2.422256E-02 6.148842E+00 1.461592E-01
75 1.031844E-02 6.148842E+00 1.447960E-01
76 7.840121E-03 6.148842E+00 1.437109E-01
77 8.056942E-03 6.148843E+00 1.426344E-01
78 9.790376E-03 6.148843E+00 1.414219E-01
79 2.868933E-02 6.148843E+00 1.379589E-01
80 1.958513E-02 6.148844E+00 1.357188E-01
81 2.634544E-02 6.148842E+00 1.325607E-01
82 6.357364E-03 6.148835E+00 1.330242E-01
83 2.592034E-02 6.148844E+00 1.299277E-01
84 7.174550E-03 6.148844E+00 1.307198E-01
85 4.660011E-03 6.148845E+00 1.301612E-01
86 1.708138E-02 6.148845E+00 1.281919E-01
87 2.339758E-02 6.148845E+00 1.255578E-01
88 2.286780E-02 6.148846E+00 1.230810E-01
89 2.357628E-02 6.148846E+00 1.205903E-01
90 2.332949E-02 6.148846E+00 1.181633E-01
91 1.825626E-02 6.148847E+00 1.163242E-01
92 1.518087E-02 6.148847E+00 1.147921E-01
93 1.198566E-02 6.148847E+00 1.135441E-01
94 5.807831E-04 6.148847E+00 1.134765E-01
Convergence was reached in nsiter = 94

iaxis_imag : 501.64s CPU 502.00s WALL ( 1 calls)

Pade approximant of anisotropic Eliashberg equations from imaginary-axis to real-axis
Cutoff frequency wscut = 0.8000

Size of allocated memory per pool: ~= 2.1113 Gb
4 4.907039E+02 5.098375E-03

Convergence was reached for N = 4 Pade approximants

Size of allocated memory per pool: ~= 2.1113 Gb
raxis_pade : 1.95s CPU 2.12s WALL ( 1 calls)

Analytic continuation of anisotropic Eliashberg equations from imaginary-axis to real-axis

Total number of frequency points nsw = 2500
Cutoff frequency wscut = 0.8000

Size of allocated memory per pool: ~= 2.9347 Gb

Size of allocated memory per pool: ~= 3.0361 Gb

Size of allocated memory per pool: ~= 3.0353 Gb

Size of allocated memory per pool: ~= -1.8074 Gb

hlee
Posts: 364
Joined: Thu Aug 03, 2017 12:24 pm
Affiliation: The University of Texas at Austin

### Re: Problem in Anisotropic superconducting gap

Dear ilias-serifi:

Did you use the input flag of "lacon=.true."?
I would like to ask you to (1) let me know the version of EPW you are using and (2) provide the EPW input.

Sincerely,

H. Lee

hpaudya1
Posts: 120
Joined: Tue Mar 21, 2017 7:11 pm
Affiliation:

### Re: Problem in Anisotropic superconducting gap

Hi ilias-serifi,

In addition to H. lee's comment,

I see some unphysical numbers in your output files, could you please double-check these?

1. The value of lambda with the Migdal-approx. is 1.2556402 while with the Eliashberg is 6.4866515.
2. You have "Fermi window (eV) = 1.0000000000E+10", can you please check why did you use this large value? and the "Electron smearing (eV) = 5.0000000000E-01"?

Best,
Hari

ilias-serifi
Posts: 59
Joined: Mon Dec 17, 2018 12:43 pm
Affiliation:

### Re: Problem in Anisotropic superconducting gap

Hello dear experts
Thanks you for the reply, i use the 5.3.1 epw version
All is good in wannierization and projections here attached the epw input
--
&inputepw

outdir = './'

iverbosity = 2

ep_coupling = .true.
elph = .true.
epbwrite = .true.

epwwrite = .true.

nbndsub = 12

wannierize = .true.
num_iter = 1000
iprint = 2
dis_win_max = 1000
dis_win_min = -10
dis_froz_max= -3.5
dis_froz_min= -5
eliashberg = .true.

proj(1) = ' :pz'
proj(2) = ': px,py,pz'

fermi_energy = -3.5501

! bands_skipped = 'exclude_bands = 1-2'

elecselfen = .true.
phonselfen = .true.
a2f = .true.

! wannier_plot_list = '2-3, 5-6, 9-16, 18'
wdata(1) = 'dis_num_iter = 5000'
wdata(2) = 'dis_mix_ratio = 0.9'
wdata(3) = 'guiding_centres = .true.'
wdata(4) = 'bands_plot = true'
wdata(5) = 'bands_num_points = 40'
wdata(6) = 'begin kpoint_path'
wdata(7) = 'G 0.00 0.00 0.00 Y 0.00 0.50 0.00'
wdata(8) = 'Y 0.00 0.50 0.00 T 0.50 0.50 0.00'
wdata(9) = 'T 0.50 0.50 0.00 Z 0.50 0.00 0.00'
wdata(10) = 'Z 0.50 0.00 0.00 G 0.00 0.00 0.00'
wdata(11) = 'end kpoint_path'
wdata(12) = 'num_print_cycles = 50'
fermi_plot = .true.
iverbosity = 0
system_2d = .true.

! fsthick = 3 ! eV
degaussw = 0.5 ! eV
conv_thr_iaxis = 1.d-02

ephwrite = .true.
kmaps = .false.
Laniso = .true.
limag = .true.
lacon = .true.
temps = 1 2 3

nsiter = 500
max_memlt = 4d0
elecselfen = .true.
phonselfen = .true.
a2f = .true.
wscut = 0.8934
! fsthick = 3 ! eV
dvscf_dir = '../phonon/save/'
filukk = './B3S.ukk'
! filqf = 'path.dat'
! 'Al_band.kpt'

nkf1 = 30
nkf2 = 30
nkf3 = 1

nqf1 = 15
nqf2 = 15
nqf3 = 1

nk1 = 10
nk2 = 10
nk3 = 1

nq1 = 5
nq2 = 5
nq3 = 1
/
13 cartesian

0.000000000 0.000000000 0.000000000
0.001118755 0.232323006 0.000000000
0.002237510 0.464646013 0.000000000
0.199995077 -0.001503953 0.000000000
0.201113832 0.230819053 0.000000000
0.202232587 0.463142060 0.000000000
0.197757567 -0.466149966 0.000000000
0.198876322 -0.233826960 0.000000000
0.399990153 -0.003007906 0.000000000
0.401108908 0.229315100 0.000000000
0.402227663 0.461638107 0.000000000
0.397752643 -0.467653919 0.000000000
0.398871398 -0.235330913 0.000000000
Best regard

hlee
Posts: 364
Joined: Thu Aug 03, 2017 12:24 pm
Affiliation: The University of Texas at Austin

### Re: Problem in Anisotropic superconducting gap

Dear ilias-serifi:
lacon = .true.
If you want to use the input flag of lacon, you have to use the most recent version of EPW v.5.4.1 included in QE v7.0 since all previous versions of EPW have a bug which I fixed in EPW v.5.4.1.

Sincerely,

H. Lee

rishab
Posts: 1
Joined: Fri Jan 28, 2022 9:56 am

### Re: Problem in Anisotropic superconducting gap

This is great, thanks for sharing.