[Wien] degeneracy problem in dipole matrix elements

[Little Grass]

Dear WIEN2k users,

I'm trying to study the dipole matrix elements of MgO at Gamma point, but
the dipole matrix elements between the degenerated bands seem to have been
mixed together. Are there any ways to sort out the degeneracy, so that the
dipole matrix elements are non-zero only for bands with the same symmetry?


The band structure can be generated from the spaghetti program. In
addition, bands with different symmetries are colored differently with
irrep.  For example, Fig1 shows the band structure from X to Gamma to X.
According to symmetry, only the bands with the same color will be dipole
allowed in px. So we expect that the dipole matrix elements (in file
mommat2) for px is zero between bands with different colors. For example,
the dipole matrix elements between band 6 and band 9 is non-zero, but the
dipole matrix elements between band 7 and 9 or 8 and 9 are zero. But if one
looks at the px component of the dipole matrix element from the optic
program, we see that at Gamma point, the dipole matrix elements between the
degenerated bands are mixed together. For example, dipole matrix elements
between band 6 and 9, between 7 and 9, and between 8 and 9 are all
non-zero. In other words, the dipole matrix is much denser than it should
be.
So is there a way to fix the dipole matrix so that the dipole matrix become
very sparse and only bands with the same symmetry have non-zero dipole
matrix elements?

Fig1:http://imgur.com/a/c7p0u

Thanks in advance for the help.
Best,
Mengxi Wu
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20161113/4a958da1/attachment.html>