Dear Dubbeldam,

Thanks for your reply. It seems I misused the two effects, i.e. the boundary effect and the finite-size effect.

In the grand-canonical approach, the intermolecular force between the substrate of interest and the image substrate can be minimized by creating the free space in the z-direction. In other words, the finite-size effect can be handled.

However, for the Gibbs approach, if the simulation box of the adsorbed phase is enlarged to avoid the finite-size effect, a fluid phase seems to appear in this simulation box.

So, is it practical to set the adsorbed phase box with a sufficiently large box volume?

Thanks in advance

Sincerely,

Jianbo Wang