[Wannier] Orthonormality of the Wannier functions

彭冠豪 bnm852i at gmail.com
Fri Mar 9 03:08:42 CET 2018

 Dear Wannier developers

I have used a combination of VASP (PAW-GGA package) and Wannier90 to obtain
MLWFs in the system of monolayer MoS2.

The band structure calculated by the TB-Hamiltonian matrix (given by
"seedname_hr.dat") can well reproduce the DFT-band structure.

However, I failed to reproduce the orthonormality of the Wannier functions
<w_m|w_n>=delta_(mn) by using the MLWFs from "seedname_w.xsf" (User Guide,
p.79, 8.15), where |w_m> and |w_n> represent the MLWFs .

For example,  <w_5|w_11>=3.79 (it should be zero)  and <w_1|w_1>= 288 (it
should be 1) in my calculation.

Q1: Could anyone give me a reason for getting these non-orthonormal
results? The grids in my numerical integration method are identical to the
grids provided by "seedname_w.xsf".

Q2: Is it reasonable that I normalize the MLWFs by myself? Since I want to
further calculate the integral <w_m|exp(iG.r)|w_n>,  where G is the
reciprocal lattice vector and r is the position vector.

Any help would be appreciated.


Guan-Hao Peng
Ph.D. Student
Department of Electrophysics
National Chiao Tung University
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/wannier/attachments/20180309/565a9717/attachment.html>
-------------- next part --------------
Wannier mailing list
Wannier at lists.quantum-espresso.org

More information about the Wannier mailing list