[DFTB-Plus-User] Charge in dftb+ calculations

[Mat Toliday]

Anshu,
If the molecules are a reasonable distance from each other (much more than the sum of the two closest atoms covalent radii) then I doubt you would see charge transfer using DFTB. The tight-binding nature of DFTB assumes that the electrons stay close to their respective nuclei. DFTB will not produce van der Walls forces by itself, you need to use the D3 package to provide any London dispersion based forces that occur, because it is a correlation effect. Correlation usually requires Moller-Plesset perturbation theory, coupled-cluster or some other form of post-Hartree-Fock type calculation.
I hope this helps,
Mat




On Monday, 16 March 2020, 06:47:13 CET, Anshu Gaur via DFTB-Plus-User <dftb-plus-user at mailman.zfn.uni-bremen.de> wrote: 





Dear Mat,

Thanks for your inputs. We are trying to see interaction between two
different molecules (charge-transfer from one to another, van-der Wals
forces and any other insteraction) and effect of these on individual
molecules. We see charge transfer from one molecule to another (differnt
charge on different atoms of molecules). When we try to see the effect
of this charge on individual molecule, the SCC cycle redistributes the
charges to all atoms. 

Regards,
Anshu

Hello Anshu,DFTB, like other electronic structure methods, tries to
determine the behavior of the electrons in the molecule you provide.
Once the approximate electron wavefunctions are determined, then the
effective charges on the atoms are calculated. This is the reason you
can only provide initial charges, which is useful when trying to get
DFTB to converge on a certain wavefunction (like a ferromagnetic or
antiferromagnetic state). Otherwise DFTB will always try to converge on
a physically reasonable ground state which will generally not be what a
user provides in terms of initial charges. Perhaps you can explain why
you were hoping to do this, it might others  help answer you question
more fully.Mat
    On Thursday, 27 February 2020, 17:52:30 CET, Anshu Gaur via
DFTB-Plus-User <dftb-plus-user at mailman.zfn.uni-bremen.de> wrote:  

Dear all,

I am trying to add charge to a molecule (different amount of charge on 
different atoms of same type), which can be done by using 
"InitialCharges" option in the Hamiltonian. However, this only 
applicable if I do SCC calculation. SCC routine re-distributes charges 
equally among all atoms if the Charge (total charge in the system) is 
provided.

Is there a way to include Initia -charges and do non-SCC calculations so 
that charges added on different atoms remain as defined during 
calculations (such as force on atoms or other properties)?

Can MaxSCCIterations be set to '0' to avoid re-distribution of charges, 
or SCCTolerance be set to an arbitrarily high value so that SCC 
converges without any re-distribution of charges?

Regards,

Anshu


_______________________________________________
DFTB-Plus-User mailing list
DFTB-Plus-User at mailman.zfn.uni-bremen.de
https://mailman.zfn.uni-bremen.de/cgi-bin/mailman/listinfo/dftb-plus-user