Question regarding free surfaces

Dear all,

I have cleaved the bulk structure of a solid (with space group of F43m) along a particular direction, and a (non-polar) surface with the thickness of about 26 A has been created. When this slab is topped with vacuum (about 100A) and boundary conditions are applied in all three directions, do we have one free surface in such a system (the top layer with under-coordinated atoms) or the bottom layer should be considered as a free surface as well? When studying the energy profile of the layers along the normal direction (by dividing the model - which had undergone a 500 ps NPT run, followed by a 200 ps NVT run- to parallel slabs with 2A thickness and evaluating the energy of each slab) I noticed the energy of the bottom layer is approximately the same as the bulk region, while as expected, the energy of the top layer is a way smaller (negative) value. I wonder whether the bottom layer should be considered as a free surface as well, and if yes, then why its behavior is different from the top layer and its energy is similar to the bulk region? Does it have anything to do with their freedom to move across the boundary along the z direction? When tracing their spacial position, it is observed that the z coordinate of all the bottom layer atoms is 0 at the beginning, but throughout the dynamic runs, it takes negative values for almost 80% of them.
I would truly appreciate your comments on this.

Best Regards,