Error running psi4 on charged molecules

Question

I am running a simple energy calculation on a conformer of 3-Sulfopropyl methacrylate, which is given my C=C(C)C(=O)OCCCS(=O)(=O)[O-].

This is my workflow:

1. Generate conformers using genetic algorithm in OpenBabel

obabel -:"C=C(C)C(=O)OCCCS(=O)(=O)[O-]" -O SPMA.pdb --gen3D -d
obabel SPMA.pdb -O SPMA_op.xyz --conformer --nconf 300 --score rmsd --writeconformers

2. Extract coordinates of conformer, and put them in psi4 input file given by:

memory 8 GB
molecule MA0 {
0 1
  C          3.89733        0.15289       -1.35762
  C          4.59280        1.05052       -0.63907
  C          5.20105        2.27534       -1.24308
  C          4.74095        0.79591        0.82319
  O          5.01564        1.66100        1.64197
  O          4.54800       -0.52000        1.09100
  C          5.28200       -1.03600        2.21300
  C          6.78400       -0.89700        2.00000
  C          7.24307       -1.67225        0.77007
  S          6.44265       -1.12891       -0.72122
  O          6.80895        0.28345       -0.80582
  O          7.02056       -1.98627       -1.75252
  O          5.02989       -1.37699       -0.43672
  H          3.77090        0.29082       -2.37251
  H          3.48966       -0.67439       -0.89456
  H          4.99668        2.29431       -2.29321
  H          6.25936        2.26265       -1.08586
  H          4.78246        3.14570       -0.78247
  H          5.04197       -2.07085        2.34094
  H          5.00988       -0.47515        3.08265
  H          7.29461       -1.27505        2.86096
  H          7.00956        0.13820        1.85038
  H          7.02443       -2.70921        0.91770
  H          8.29303       -1.50198        0.65390
  H          7.75698        0.36077       -0.99601
  H          7.95459       -1.75656       -1.87793
}

set {
  scf_type          = df
  freeze_core       = true
 
}
E, wfn = energy('HF-3c', return_wfn=True)

oeprop(wfn, 'DIPOLE', 'QUADRUPOLE', 'MULTIPOLE(4)', 'MO_EXTENTS', 'MULLIKEN_CHARGES', 'LOWDIN_CHARGES', 'WIBERG_LOWDIN_INDICES', 'MAYER_INDICES', 'NO_OCCUPATIONS', title='Methacrylic acid (MA0)')

print_variables()

3. Now I run psi4 on this file:

psi4 -n 2 -i SPMA_conf.in -o SPMA.out

When I run this, however, I get a wall of error messages.

Printing out the relevant lines from the Psithon --> Python processed input file:
      H          7.00956        0.13820        1.85038
      H          7.02443       -2.70921        0.91770
      H          8.29303       -1.50198        0.65390
      H          7.75698        0.36077       -0.99601
      H          7.95459       -1.75656       -1.87793
--> ""","MA0")
    core.IO.set_default_namespace("MA0")
    core.set_global_option("SCF_TYPE", "df")
    core.set_global_option("FREEZE_CORE", "true")
    E, wfn = energy('HF-3c', return_wfn=True)
    oeprop(wfn, 'DIPOLE', 'QUADRUPOLE', 'MULTIPOLE(4)', 'MO_EXTENTS', 'MULLIKEN_CHARGES', 'LOWDIN_CHARGES', 'WIBERG_LOWDIN_INDICES', 'MAYER_INDICES', 'NO_OCCUPATIONS', title='Methacrylic acid (MA0)')

This seems to be the relevant portion of the error message...

I have run the very same process on other conformers of other non-charged molecules, but I have not run into this problem. Is it because 3-Sulfopropyl methacrylate is charged?

A similar problem arises when I run another charged molecule, like C=C(C)C(=O)OCC[N+](C)(C)C.

How can I fix this problem? Any advice you have would be appreciated!

Answer 1

As you already figured out, the error is that the charge and multiplicity are inconsistent with the molecule; this is what Psi4 tells you:

qcelemental.exceptions.ValidationError: Inconsistent or unspecified chg/mult: sys chg: None, frag chg: [0.0], sys mult: None, frag mult: [1]

The problem is that the system has 111 electrons, which is an odd number, so you can't construct a spin-singlet wave function. Valid multiplicities for the neutral molecule are 2 (doublet), 4 (quartet), 6 (sextet), etc. The anion and cation both have an even number of electrons, so they can be in a singlet, triplet, quintet, etc.

Note also that freeze_core does nothing for self-consistent field calculations in Psi4.