Convention of electric field direction in VASP
Posted: Thu Jan 19, 2023 7:13 am
Dear all,
Sorry to bring up a topic that has mentioned several times in the forum - but I have not been able to find a satisfactory conclusion so far.
It is known that EFIELD corresponds to the magnitude of the electric force field, F = qE where q is a test charge and E is the electric field, see for instance forum/viewtopic.php?f=3&t=17653&p=18816 ... ial#p18816
I am confused about the direction of the electric field in VASP. By definition E = -\grad V where V is the electric potential. So, if LOCPOT (plotted without V_xc) gives the sawtooth potential to have a positive gradient along the positive z direction, I expect to get a electric field pointing in the negative z direction.
For example, if I compute the LOCPOT of graphene with EFIELD = +1.0 eV/Ang, plotting the planar averaged electrostatic potential from LOCPOT
gives me a negative gradient along the positive z direction, which I think corresponds to an electric field of 1.0 V/Ang pointing in the positive z direction. I would then expect an accumulation (depletion) of electron density below (above) the graphene layer. However, if I compare with the case without electric field (EFIELD = 0) by plotting the charge density difference \rho_diff = \rho_field - \rho_nofield
I see the reverse behavior where the electron density is accumulated (depleted) above (below) the graphene layer, which suggests to me that the electric field is actually pointing in the negative z direction, to be consistent with fundamental principles that electrons travel in opposite direction of the electric field.
Wondering if I am missing something here. Appreciate also clarification the sign of the test charge q in F = qE. Thank you so much for your attention!
Please see attached zip folder for the inputs I used for this calculation.
Best,
Nicholas
Department of Physics
National University of Singapore
Sorry to bring up a topic that has mentioned several times in the forum - but I have not been able to find a satisfactory conclusion so far.
It is known that EFIELD corresponds to the magnitude of the electric force field, F = qE where q is a test charge and E is the electric field, see for instance forum/viewtopic.php?f=3&t=17653&p=18816 ... ial#p18816
I am confused about the direction of the electric field in VASP. By definition E = -\grad V where V is the electric potential. So, if LOCPOT (plotted without V_xc) gives the sawtooth potential to have a positive gradient along the positive z direction, I expect to get a electric field pointing in the negative z direction.
For example, if I compute the LOCPOT of graphene with EFIELD = +1.0 eV/Ang, plotting the planar averaged electrostatic potential from LOCPOT
gives me a negative gradient along the positive z direction, which I think corresponds to an electric field of 1.0 V/Ang pointing in the positive z direction. I would then expect an accumulation (depletion) of electron density below (above) the graphene layer. However, if I compare with the case without electric field (EFIELD = 0) by plotting the charge density difference \rho_diff = \rho_field - \rho_nofield
I see the reverse behavior where the electron density is accumulated (depleted) above (below) the graphene layer, which suggests to me that the electric field is actually pointing in the negative z direction, to be consistent with fundamental principles that electrons travel in opposite direction of the electric field.
Wondering if I am missing something here. Appreciate also clarification the sign of the test charge q in F = qE. Thank you so much for your attention!
Please see attached zip folder for the inputs I used for this calculation.
Best,
Nicholas
Department of Physics
National University of Singapore