slab calculation with MBJ potential

[zj]

Dear all,

I am now doing a slab calculation. It is OK if the LDA or GGA is used.
However, I want to improve the band gap, then the MBJ potential
is probably the best choice. Hybrid functional is not possible for my
large system.

As manual said, "MBJ calculations tend to diverge for surface calculations".
I indeed found the slab's total energy diverged in my calculation,
even though the bulk's MBJ calculation can be converged.

I wonder does anybody have the experience to do slab calculation with MBJ?

In the wien2k, there is a trick to calculate the slab with MBJ.
It seems that wien2k use the bulk's [avrg grad rho /rho] for the slab
to get the converged result. However, I have not checked it at present.
Does vasp can do the similar things?

Thanks

[admin]

In the vacuum layer the electron density and the kinetic energy density drops to zero and this causes that the total energy diverge. MBJ is not applicable to surfaces and layers in vasp.

[zj]

Dear admin,

Thank you very much for your reply.