search

openmopac / mopac

Molecular Orbital PACkage
http://openmopac.net
GNU Lesser General Public License v3.0
111 stars 31 forks source link

Bug in BZ with symmetry operations #190

Closed godotalgorithm closed 6 months ago

godotalgorithm commented 6 months ago

Discussed in https://github.com/openmopac/mopac/discussions/187

Originally posted by **kamischi** December 20, 2023 I perform a PM7 calculation on a 2D periodic honeycomb system of Sc2O3. The actual calculation runs nicely without any obvious faults. But when I try to get the band structure with "bz", I get this error: Secular determinant generated by symmetry operations may be faulty. As suggested from the output I tried the options SYMTRZ and ROTFOK in bz. This produced some more output, but it did not really give me a better idea of what is wrong. I have already tried NOSYM as keyword in mopac and switched from internal coordinates to xyz, but none of this made any change. So far, I did quite a number of similar calculations with a 2D honeycomb system with Ti2O3 and never had this problem. Does anyone have any ideas? Input/Output files are appended as zip archive. Best wishes - Michael. [Sc2O3 6-Ring.zip](https://github.com/openmopac/mopac/files/13729446/Sc2O3.6-Ring.zip)
godotalgorithm commented 6 months ago

Now that BZ isn't crashing because of an Intel compiler bug, I have tested this fix and found that it isn't completely working yet. The assignment of symmetry-equivalent atoms does seem to be fixed, but something else is still triggering the error and the assignment of unit cells for the symmetry-equivalent atoms still looks weird.

godotalgorithm commented 6 months ago

Finally fixed by #193.