Dear Researchers,

I am performing a simulation regarding the diffusion of water molecules in an MFI zeolite structure. The parameters are exactly obtained from a high-quality paper. I first did a 300 ps equilibration and a 2 ns production run using Gromacs. However, after 2 ns and even longer simulation times, the diffusion coefficient is far from the reported one at small water contents (but diffusion coefficient numbers are similar at high water content). On the other hand, the diffusion numbers are very close to the calculated diffusion coefficient at the equilibration level (200 ps) where water molecules are separated to some extent. As you know, any analysis at the equilibration level is not reliable.

I have two questions!

1- In the case of water in zeolite, how much the simulation time is important to get a reliable diffusion coefficient? Because, as I saw at longer simulation times most of the water molecules (at low water contents) form one big cluster in the zeolite pores leading to a decrease in diffusion coefficient due to forming hydrogen bonds).

2- Is there any specific consideration for calculating the diffusion coefficient of water at small water contents(Small hydration levels)? I think we have to consider some strategies for low water content (e.g., smaller simulation times, ...). Is it true?

Thanks alot