Discrepancy in data for mvc-11541

Could you please verify the point group of mvc-11541? As far as I know a simple hexagonal crystal structure cannot have “mmm” point group symmetry.

Our reported crystal system and point group symbol is a function of the tolerances we use for symmetry finding, so that may be part of the issue here. There may also be a bug in the spglib library that we use under the hood to return this information, in which case I encourage you to raise this issue e.g. on their mailing list so that this gets fixed upstream for all users of the library.

The MP website uses a 0.1 Å length tolerance and a 5 degree angle tolerance (the symprec and angle_tolerance paramers for pymatgen’s SpacegroupAnalyzer, respectively.

For example, using our API:

from pymatgen import MPRester
from pymatgen.analysis.structure_analyzer import SpacegroupAnalyzer

m = MPRester()
s = m.get_structure_by_material_id("mvc-11541")
spa = SpacegroupAnalyzer(s, symprec=0.1) # default angle_tolerance is 5
print(spa.get_crystal_system()) # prints "hexagonal"
print(spa.get_point_group_symbol()) # prints "mmm"

versus

spa = SpacegroupAnalyzer(s, symprec=0.01) # stricter tolerance
print(spa.get_crystal_system()) # prints "monoclinic"
print(spa.get_point_group_symbol()) # prints "2/m"
1 Like