Orbitals are losing symmetry during optimization

[ktanovitz]

Hi,

I tried to optimize the ground state for propargyl bromide (C3H3Br), but I noticed that some of the orbitals were losing symmetry during the optimization. I am seeing this error:

Orbital Energies (a.u.) and Symmetries

Warning : Irrep of orbital( 3) could not be determined
Warning : Irrep of orbital( 5) could not be determined
Warning : Irrep of orbital( 10) could not be determined
Warning : Irrep of orbital( 12) could not be determined
Warning : Irrep of orbital( 13) could not be determined
Warning : Irrep of orbital( 14) could not be determined
Warning : Irrep of orbital( 15) could not be determined
Warning : Irrep of orbital( 16) could not be determined
Warning : Irrep of orbital( 17) could not be determined
Warning : Irrep of orbital( 23) could not be determined
Warning : Irrep of orbital( 24) could not be determined
Warning : Irrep of orbital( 26) could not be determined
Warning : Irrep of orbital( 27) could not be determined
Warning : Irrep of orbital( 29) could not be determined
Warning : Irrep of orbital( 30) could not be determined
Warning : Irrep of orbital( 31) could not be determined
Warning : Irrep of orbital( 34) could not be determined
Warning : Irrep of orbital( 38) could not be determined
Warning : Irrep of orbital( 39) could not be determined
Warning : Irrep of orbital( 42) could not be determined
Warning : Irrep of orbital( 43) could not be determined
Warning : Irrep of orbital( 44) could not be determined
Warning : Irrep of orbital( 45) could not be determined
Warning : Irrep of orbital( 46) could not be determined
Warning : Irrep of orbital( 47) could not be determined
Warning : Irrep of orbital( 48) could not be determined
Warning : Irrep of orbital( 49) could not be determined
Warning : Irrep of orbital( 51) could not be determined
Warning : Irrep of orbital( 53) could not be determined
Warning : Irrep of orbital( 55) could not be determined
Warning : Irrep of orbital( 56) could not be determined
Warning : Irrep of orbital( 59) could not be determined
Warning : Irrep of orbital( 60) could not be determined
Warning : Irrep of orbital( 62) could not be determined
Warning : Irrep of orbital( 63) could not be determined
Warning : Irrep of orbital( 65) could not be determined
Warning : Irrep of orbital( 66) could not be determined
Warning : Irrep of orbital( 67) could not be determined
Warning : Irrep of orbital( 68) could not be determined
Warning : Irrep of orbital( 69) could not be determined
Warning : Irrep of orbital( 71) could not be determined
Warning : Irrep of orbital( 72) could not be determined
Warning : Irrep of orbital( 75) could not be determined
Warning : Irrep of orbital( 80) could not be determined
Warning : Irrep of orbital( 81) could not be determined
Warning : Irrep of orbital( 82) could not be determined
Warning : Irrep of orbital( 83) could not be determined
Warning : Irrep of orbital( 84) could not be determined

Here is the geometry and the rem section:

Does anyone have any suggestions on how the orbital symmetry can be preserved?
-Kyle

[jherbert]

(a) Does Q-Chem recognize the point-group symmetry of your input structure? (Should be printed just below the coordinates in Standard Nuclear Orientation).

(b) Please post an input file that I can copy/paste, not a screenshot. Use the formatted-text button </>

[ktanovitz]

a. Qchem does recognize the point-group symmetry of the input structure.
b.`$molecule
0 1
C 0.0000000 0.5523422 -2.8521583
C 0.0000000 -0.1531284 -1.8742547
C 0.0000000 -1.0101064 -0.6874045
H 0.0000000 1.1712459 -3.7106485
H 0.0000000 -1.9606782 -1.2600268
H 0.0000000 -1.9425407 -0.0864968
Br 0.0000000 0.1262903 0.8775867
$end

$rem
BASIS = 6-31G*
JOB_TYPE = Optimization
METHOD = wB97XD
MEM_TOTAL = 10000
$end`

[jherbert]

Not sure what the problem is, it persists even with THRESH = 14 and I notice that if the final, optimized geometry is used in a new calculation, Q-Chem doesn’t recognize the symmetry unless the tolerance is loosened (e.g., SYM_TOL = 2). Seems like the optimizer is gradually losing point group symmetry. It seems to be related to the new optimizer (I am using Q-Chem 6.1), because the problem goes away if I force the use of the old optimizer with GEOM_OPT_DRIVER = OPTIMIZE. I will submit a bug ticket. For now, use of the old optimizer is a workaround.

UPDATE: the problem is that you have provided an unphysical (extremely distorted) starting geometry. The new optimizer does not attempt to preserve symmetry, and optimizes to something that is no longer planar (and seems reasonably to my eye). The old optimizer will preserve point-group symmetry, but given your starting geometry that means that it dissociates into a planar vdW complex involving broken bonds to the H atoms. I don’t think this is a bug, just an artifact of an unreasonable starting point.