Error: Problem with modes file

[srp_001]

I am trying to compute scattering rates and mobility in 2D graphene, but I am running into an error that I am not able to fix:
elphon_shuffle_wrap : error # 1: Problem with modes file

I have looked at a previous post on this error and have made sure that the nqs specified in EPW.in are consistent with nscf.in file.

I have copied my epw.in file below:

Code: Select all

&inputepw
prefix = '2D'
amass(1) = 12.0107
outdir = './'

elph = .true.
epbwrite = .true.
epbread = .false.
ephwrite = .true.

epwwrite = .true.
epwread = .false.

lifc = .false.
carrier = .false.
scattering = .true.
scattering_serta=.true.
int_mob = .true.

ep_coupling = .true.
elecselfen = .true.
phonselfen = .true.

wannierize = .true.
dis_froz_min= -1d3
dis_froz_max= -0.9d3
dis_win_min = -1d3
dis_win_max = 1d3
num_iter = 300
iprint = 2
iverbosity = 1

efermi_read = .true.

a2f = .true.

fsthick = 2.80284905 ! eV
temps = 300! K
degaussw = 0.05 ! eV

wdata(1) = 'bands_plot = .true.'
wdata(2) = 'begin kpoint_path'
wdata(3) = 'G 0.0000000000 0.0000000000 0.0000000000 M 0.0000000000 -0.5000000000 0.0000000000'
wdata(4) = 'M 0.0000000000 -0.5000000000 0.0000000000 K 0.3333333333 -0.6666666667 0.0000000000'
wdata(5) = 'K 0.3333333333 -0.6666666667 0.0000000000 G 0.0000000000 0.0000000000 0.0000000000'
wdata(6) = 'end kpoint_path'
wdata(7) = 'num_print_cycles = 1'
wdata(8) = 'bands_plot_format = gnuplot'
wdata(9) = 'guiding_centres = .true.'
wdata(10) = 'dis_num_iter = 300'
wdata(11) = 'dis_mix_ratio = 1.0'

proj(1) = 'random'
dvscf_dir = './save'
filukk = './C.ukk'

nk1 = 12
nk2 = 12
nk3 = 1

nq1 = 12 
nq2 = 12
nq3 = 1

nqf1 = 20
nqf2 = 20
nqf3 = 20

nkf1 = 20
nkf2 = 20
nkf3 = 20
/

I have copied the last 50 or so lines of epw.out below:

Code: Select all

         140 of  144 on ionode
          141 of  144 on ionode
          142 of  144 on ionode
          143 of  144 on ionode
          144 of  144 on ionode
     MMN calculated

     Running Wannier90

     Wannier Function centers (cartesian, alat) and spreads (ang):

     (   0.71046   0.28505   0.86047) :  10.94706
     (   0.84235  -0.00314   2.21271) :   2.06681
     (   0.27560   1.16504  10.20454) :  16.25189
     (   0.44127   0.32934   2.60448) :  10.87585
     (   0.05324   0.74945   2.32366) :  10.35836
     (   0.50875   0.85297   2.16535) :   8.81059
     (   0.41970   0.38550   1.88425) :  14.49698
     (  -0.48503   0.41892  10.02872) :   7.02151

     -------------------------------------------------------------------
     WANNIER      :    120.87s CPU    120.96s WALL (       1 calls)
     -------------------------------------------------------------------

     Calculating kgmap

     Progress kgmap: ########################################
     kmaps        :      0.08s CPU      0.09s WALL (       1 calls)
     Estimated size of gmap: ngxx = 7569
     Symmetries of Bravais lattice:  24
     Symmetries of crystal:          24

 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
     Error in routine elphon_shuffle_wrap (1):
      Problem with modes file
 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

     stopping ...

Any suggestions regarding how I should fix this would be greatly appreciated.
Thank you!

[sponce]

Hello,

This means that there is an incompatibility between your phonon calculation (dvscf files) and the current EPW calculations.

Note that the new version of EPW v5.8 (to be released very soon) will include support for 2D materials.

You can already download the development version here if you are interested https://gitlab.com/epw/q-e

Best wishes,
Samuel

[adenchfi]

I also had this issue. I made sure to use QE v 7.2 for the scf+nscf+ph+epw calculations. The nq grid is the same between ph/EPW.
I have also checked that the patterns files exist, and they appear to be of the correct format for the read_disp_pattern usage. Using QE v 7.1 rerunning all the calculations gives the same result.

It appears EPW does not like dvscf_dir to be an absolute file path, at least on my cluster. Replacing the full dvscf_dir filepath with '../save/' entirely fixes the issue. I spent entirely too long figuring this out...