On Tue, 3 Apr 2007, Rob wrote:
I feel I can see one band that is approximately flat around normalized frequency 0.28, but then there are some other bands in there, as well. Is my problem due to a too-small supercell or should these other bands actually be in the gap as well?
Yes, you should see lots of bands in the gap--more and more as you increase the supercell size.
The reason is that a slab does not have a complete band gap, when you include the light cone. And when you create a point defect, you break translational symmetry, which breaks conservation of k, and therefore all of the light cone modes fall into the "gap" in a band-structure calculation.
In an MPB calculation, the only way to distinguish a point-defect resonant mode is to look at the localization of the field. A resonant mode will have a large fraction of its field concentrated in a small region around the defect, whereas other modes will not. The compute-energy-in-objects and output-energy-in-objects functions (if I am remembering their names correctly) will help you here.
Why am I trying to use MPB rather than MEEP to find resonant band frequencies? I am new to both of these software packages and cannot understand yet how to simulate non-rectangular unit cells in MEEP (despite your posts on this on theMEEP list). But that is a discussion for a different list.
You have a fundamental misunderstanding if you think that the unit cell is at all relevant for a point-defect calculation. For a point defect, there is no translational symmetry so you might as well use a rectangular supercell.
Cordially, Steven G. Johnson _______________________________________________ mpb-discuss mailing list [email protected] http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/mpb-discuss
