[gmx-users] Dipole moment and dielectric constant - gmx dipoles

[Iman Ahmadabadi]

Dear Gromacs users,

I'm trying to reproduce the gmx dipoles results via python code. In
calculating the fluctuations of total dipole moment, which is the sum on
dipole moment of individual molecules in the box, M = \sum q_{i}*R_{i}, in
which the charges q_{i} are in electron unit and position R_{i} are in nm.
For calculating M for each time step (frame), the x,y, and z component of
each atom is multiplied to their charges, and then summing on all atoms.
The average of M is on all trajectory frames, meaning that summing on M for
all frames and then dividing by the number of frames. Here, the acquired M
is divided by proper constant (0.020819434 e
<https://en.wikipedia.org/wiki/Elementary_charge>·nm
<https://en.wikipedia.org/wiki/Nanometre>) to set it in the Debye unit.

For M^{2}, M.M (dot product of M by M) is calculated for each frame and
then summing on all frames, followed by dividing by the number of frames.
But the results are larger than gmx dipoles, I wanted to ask that there is
something gmx dipoles is using in order to calculate the <M> and <M^2>,
causing this difference? I appreciate any help in this regard.

Best regards,
Iman