Query on electrostatic treatment for slab/interface systems (2D periodicity)

User discussions
1

GROMACS version:
GROMACS modification: Yes/No
Here post your question
Dear GROMACS developers and users,

I have a question regarding the electrostatic calculation for systems with 2D periodicity (periodic in XY, non-periodic in Z), such as membranes or liquid-vapor interfaces.

Specifically, I would like to clarify the method used in GROMACS:

  1. Is the standard approach for such systems the one where we add a vacuum layer in the Z-direction and apply the standard 3D Ewald/PME method, or does GROMACS implement a dedicated 2D Ewald/PME algorithm (like Ewald2D/PME2D) for this geometry?

  2. I understand that setting ewald-geometry = 3dc in the .mdp file introduces a correction. Is this correction an implementation of the 2D Ewald sum, or is it a correction applied on top of a 3D Ewald summation with a vacuum layer to effectively mimic a 2D electrostatic environment?

  3. If GROMACS does have a native implementation of the true 2D Ewald/PME method, could you please point me to the specific source code files or directories where it is implemented?

Thank you for your time and guidance.

2

There no 2D-Ewald implementation in GROMACS. The 3dc geometry works as in the manuscript referred in the manual. This simply/only cancels the dipole-dipole interaction term along Z. Thus the dipole-quadrupole term is the largest left and the dipole is much smaller with this correction, such that periodic interactions are very small when adding some vacuum.