[gpaw-users] NEGF-STM

Lars Smykalla lars.smykalla at physik.tu-chemnitz.de
Fri Dec 2 18:58:05 CET 2011


Hi,

I read a localized basis is necessary for STM simulation, so from the  
tutorial and documentation I it seems only possible for lcao mode.
But now I read it also works when I create Wannier functions from a plane  
wave or grid based calculation.

What python code do I have to use exactly to do this and to dump in this  
case the overlap matrix and the Hamiltonian matrix?



Further, I don't understand in the NEGF-STM tutorial:

# Here the keyword 'cvl' refers  to the number of basis functions in the
# in the convergence layer, i.e. for the present system four atomic layers  
are used.
dump_hs(calc, 'srf', region='surface', cvl=4*4*9)


There a 16 Ag atoms in 4 layers, but where does this 9 come from?


Greetings,
Lars


More information about the gpaw-users mailing list