Page 1 of 1

Error in routine read_control_ph (1): wrong elph

Posted: Tue Jul 09, 2024 1:42 am
by bhawna
Hi, I am trying to run a direct evaluation of electron-phonon matrix elements along the
selected high-symmetry lines using density-functional perturbation theory but I am getting this error:
Error in routine read_control_ph (1): wrong elph
My ephline.in file is :
&inputph
prefix = 'XYZ'
fildvscf = 'dvscf'
ldisp = .true.
recover = .true.
fildyn = 'XYZ.dyn'
tr2_ph = 1.0d-16
qplot = .true.
q_in_band_form = .true.
electron_phonon = 'prt'
kx = 0.5
ky = 0.5
kz = 0.5
/
3
0.5000 0.0000 0.5000 20 # X
0.0000 0.0000 0.0000 20 # Gamma
0.5000 0.5000 0.5000 1 # L
I am using this command:
module load QuantumESPRESSO/7.3
mpirun -n 48 ph.x -npools 48 -i ephline.in > ephline.out
Please let me know what might be the issue.

Re: Error in routine read_control_ph (1): wrong elph

Posted: Wed Jul 10, 2024 12:33 am
by stiwari
Hi Bhawna,

The problem is with the tag electron_phonon = 'prt'. Please check the Phonon code input page https://www.quantum-espresso.org/Doc/INPUT_PH.html#idm226 for the possible options for electron-phonon tag.

Best regards,
Sabya.

Re: Error in routine read_control_ph (1): wrong elph

Posted: Wed Jul 10, 2024 2:11 pm
by bhawna
Hi,
Thanks for your reply. When I use 'interpolated' tag for electron-phonon. Then, also error is coming at kx = 0.5 stating that error might be in previous line. I want to calculate the matrix elements for initial state set at "L" high symmetry point. Please let me know the proper way to perform this calculation.

Thanks,
Bhawna

Re: Error in routine read_control_ph (1): wrong elph

Posted: Sat Jul 20, 2024 3:15 am
by Shashi
Hi Bhawna,

When you specify ldisp = .true., you need to specify k-grid not a k-path.
If you want to read the "L" point coordinates, you can set ldisp = .false. and trans = .false.

regards,
Shashi

Re: Error in routine read_control_ph (1): wrong elph

Posted: Thu Jul 25, 2024 4:53 pm
by stiwari
Hi Bhawna,

One more point, you need to provide the entire k-path and not just the high-symmetry points. With that 'prt' should work.

Best,
Sabya.