How can I fix a "bi-layer" during an NPT simulation?

User discussions
1

GROMACS version: 2025.3

Dear all,
My name is Aurora Fatiga, and I work at CNR-ITM (Rende, Cosenza, Italy).
I’m having some issues with GROMACS; I’ll explain my question.

I would like to simulate, using GROMACS, a system consisting of two layers (AB) of g-C3N4 (periodic-molecules) and solvent.
This is a semi-isotropic (in the Z direction) NPT (300K, 1 bar, 1 ns) on a triclinic simulation cell with the following dimensions: (8.10000 7.01480 10.00000 0.00000 0.00000 -4.05000 0.00000 0.00000 0.00000).
After running a few simulations, I encountered a problem that I would prefer not to have:
the g-C3N4 layer at Z-min (approximately 0 nm) is shifted to Z-max due to the 3D PBC.

Is it possible to fix the g-C3N4 layers in place during the NPT simulation?

Thank you for your attention,

Aurora

2

But this shift is irrelevant for the physics. What is the issue?

3

Thank you for your reply.

For the analyses I will be doing, I would prefer that the gC3N4 stay in its initial position.

4

If your analysis cannot take into account periodic boundary conditions, the easiest solution is to move your layer up by a fraction of a nanometer.