GROMACS version:2020
GROMACS modification: No
Here post your question
Dear all,
When I simulate the freezing of saline under an external electric field, the simulation system gets compressed when the set electric field is >0.4 V/nm. How should I set up the mdp file during the simulation to achieve a normal simulation process?
I guess you are using PME. Are you aware that the electric field is enhanced by the dielectric of your solution? This will be more than two orders of magnitude for saline. See: Electric fields - GROMACS 2025.3 documentation
Yes, I used PME. I did overlook the influence of electric fields on the dielectric properties of solutions, but I also simulated the freezing of pure water under electric fields, and this problem also occurred in this system. I saw that Bug with electric field simulation using gromacs2020 (#3800) · Issues · GROMACS / GROMACS · GitLab Mentioning that this is a bug in GROMACS2020, does it have any reference value?
But water has dielectric permittivity of around 80, so the field will be around a factor 80 higher that what you specify in the mdp file.
The bug you mention only affects the 2020 release.
This strength of electric field is used in several literature I have seen, so I am trying to reproduce this strength of electric field
Yes, but is that the electric field applied as we apply it in GROMACS or the actual electric that the effect on the system corresponds to?
thanks for your suggestion. The electric field that literature mentioned is applied in GROMACS, and the GROMACS I used is 2020 version, it maybe cause the question.
The issue in 2020 was only with parallel simulations. I would not think that that bug affected their results.
yes, I use other version GROMACS do it, it also has this result.May I ask if you have any suggestions for reference?
What do you mean with “for reference”?
I just ran a simulation with an input field of 0.4 V/nm and PME and get an extreme water orientation, more than 50%. A field of 0.4/epsilon is still high, but leads to less extreme orientation.
When the system consists of neutral update groups, e.g. only water, you can set epsilon-surface=1 and you actually get the field you set in the mdp file.
Thank you for your feedback. What I mean is, could you provide an MDP file setting that simulates this system. I will install your feedback and simulate it again.
If you want an external field of 0.4 while using PME, you need to specify a field of 0.4/epsilon, where epsilon is the dielectric permittivity of your solution. You can obtain that using gmx dipoles -eps with a simulation of your system without electric field.
Thanks for your advice, I will try to do it.