Can we do independent meshes for each layer of atoms?

In our team we are studying the behavior of a crystalline surface of GaAs(111)A in the cubic"Zinc Blende" structure (that is equivalent to two fccc, one of Ga and the otherr translated by 1/4,1/4, 1/4 of As); with Abinit. Until now the fcc configuration is build with one atom and then by translations and periodicity the structure is made. But those transformations are defined only once and affects both Ga and As layers. And we asked ourselves if there is a chance to build independently two different meshes one for the Ga layer and another for the As layer.
It would be very useful for us and for our work if you could clarify our doubts. Thank you for your answers.

Hi Saelma,

If by mesh you mean real space or reciprocal space distribution of points in the simulation box, the short answer is no. The system that you define based by the vectors that you define in the input will be your simulation box which will be equally fragmented in both real and reciprocal space given your ecut, respectively your number of k-points. Ecut along your lattice parameters definition will determine the size of your FFT mesh, nonetheless ABINIT allows for a flexible definition of the FFT mesh irrespective of the ecut input (see ngfft variable).

I hope I got your question right.
Good luck!