Dear Siesta users, I've notice another peculiar behaviour of the siesta(QE). The thing is, i'm trying to model slab of finite thickness (7,9,11 monolayers) of the CH3NH3PbI3 perovskite. Everything is checked against vacuum thickness, dipole correction and k-point convergence.
The problem is: If i use symmetrical slab with PbI2 termination on both sides i get different bandgap(i know it is wrong for the DFT, but still) and surface states compared to assymetrical termination PbI2-CH3NH3I. I dont see a reason why it should be different (CH3NH3I-CH3NH3I termination does not do anything. whereas PbI2-CH3NH3I. reduces the bandgap ). If it is due to PbI2 termination i should see it in both cases, but i dont. Do i miss something ? With Best Regards, Andrei Buin.
