[DFTB-Plus-User] information of basis functions and confinement potential

张文星 zhangwenxing at tyut.edu.cn
Thu Apr 9 19:13:18 CEST 2015


Dear Gao,
You are welcome. Right, the basis functions are not available in the SK files.
LCAO method only guarantees you could use a parameterized model Hamiltonian to reproduce the band structure of solid.
But we have no idea about the exact appearance of the basis used because it is necessary only when we want to
check the physics of the basis for example we develop one tigh binding model. Otherwise, if you need to calculate some kind
of properties then just calculate it with respect to your band structure like the DFT do. If you need to derive some kind of formula
which include the basis function, then you have to transform your formula into terms of Hamiltonian matrix element or Overlap
matrix element, then you can calculate them. Or you really need the exact appearance of the basis function, then atomic orbital
is one choice, but you need to get out the relationship between the matrix elements in H and the electronic integral with respect
to the atomic orbital. Pherhaps, wannier function should be a better choice than atomic orbital. I could not to offer more help on
your question because I have not got out any basis function for tight binding model until now, and it is not necessary in my study.
best,
Zhang

---------- Origin message ----------
>From:"Gao Bin" <bin.gao at uit.no>
>To:"User list for DFTB+ related questions" <dftb-plus-user at mailman.zfn.uni-bremen.de>
>Subject:Re: [DFTB-Plus-User] information of basis functions and confinement potential
>Date:2015-04-09 21:46:45
 
Dear Zhang,
 
Thank you for your reply. Actually, I found some published atomic basis functions at http://www.dftb-plus.info/tools/waveplot/ (angular momentum, exponents and coefficients). As you may notice, different SK sets can use different atomic basis functions, that are not available in the SK files.
 
My question, if I rephrase, is that how one can get these basis functions (angular momentum, exponents and coefficients) and the confinement potential (or the wave function compression radius if harmonic potential used, see eq. (4) in J. Chem. Theory Comput. 2013, 9, 338). These are important for post-processing after the DFTB calculations.
 
Cheers
GaoFrom: dftb-plus-user-bounces at mailman.zfn.uni-bremen.de <dftb-plus-user-bounces at mailman.zfn.uni-bremen.de> on behalf of 张文星 <zhangwenxing at tyut.edu.cn>
Sent: Thursday, April 9, 2015 4:06 AM
To: User list for DFTB+ related questions
Subject: Re: [DFTB-Plus-User] information of basis functions and confinement potential Dear Gao,
The sk files include only the parameters used in the formulated Hamiltonian and Overlap matrix.
The two matrix are well formulated according to the tight binding model and the basis functions
are atom or atomlike orbitals which you can refer to the original Slater Koster paper in physical review about 60 years ago.
best wishes,
Zhang

---------- Origin message ----------
>From:"Gao Bin" <bin.gao at uit.no>
>To:"User list for DFTB+ related questions" <dftb-plus-user at mailman.zfn.uni-bremen.de>
>Subject:[DFTB-Plus-User] information of basis functions and confinement potential
>Date:2015-04-09 01:22:25
 
Dear all,
 
I have a question regarding the DFTB parameters, or the Slater-Koster files, from which we will know for instance the matrix elements of overlap and Hamiltonian.
 
But how could we know the basis functions (the coefficient, angular number, exponent, see for example eq.(3) in J. Chem. Theory Comput. 2013, 9, 338) and the confinement potential (or the wave function compression radius if harmonic potential used, see eq. (4) in J. Chem. Theory Comput. 2013, 9, 338)? Do we have to read the original literature? Or is it possible to find such information in the parameter files? Thank you in advance.
 
Cheers
Gao
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.zfn.uni-bremen.de/pipermail/dftb-plus-user/attachments/20150410/ee8449f2/attachment.html>


More information about the DFTB-Plus-User mailing list