[DFTB-Plus-User] too little or superfluous input for spin-polarised calculation

[Peter Klaver]

Hello people,

I seem to have a problem preparing the input file for a spin-polarised calculation.

Some experimental work suggested that in graphene-like defective structures, some C atoms may carry a magnetic moment. I want to see if a dftb+ calculation will show any such thing.

If I put the following lines in the Hamiltonian section of dftb_in.hsd, 
———————————————————————
SpinPolarisation = Colinear {
UnpairedElectrons = 0.0
InitialSpins = {
AtomSpin = {
Atoms = 1:7801
SpinPerAtom = 1.0
}
}
}
———————————————————————

then dftb+ says
-> Missing child: SpinConstants

If I add then add the lines
———————————————————————
SpinConstants = {
C = {
# Wss Wsp Wps Wpp
-0.028 -0.024 -0.024 -0.022
}
}
———————————————————————

then dftb+ says:
ERROR!
-> Superfluous data found.
Path: dftbplusinput/Hamiltonian/DFTB/SpinConstants/C
Line: 65-68 (File: dftb_in.hsd)

So without the SpinConstants lines there is something missing, with the SpinConstants lines there is too much.

If I only put in the line
SpinConstants = {}

then dftb+ says:
ERROR!
-> Missing child: C
Path: dftbplusinput/Hamiltonian/DFTB/SpinConstants
Line: 64-65 (File: dftb_in.hsd)

Putting in a line for C without data, like
C = {}

then dftb+ says:
ERROR!
-> Unexpected end of data
Path: dftbplusinput/Hamiltonian/DFTB/SpinConstants/C
Line: 65-65 (File: dftb_in.hsd)

So I can have too much data, of various grades of too little data.

Does anyone have an idea on how to put in the right data for a spin-polarised calculation?

Kind regards,
Peter
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.zfn.uni-bremen.de/pipermail/dftb-plus-user/attachments/20240910/a7537d60/attachment.htm>