F/I SAPT0 problem with three fragments

Dear psi4 users,

I have a question about F/I-SAPT0 computations. I recently changed version of PSI4 to 1.3 and tried to run I-SAPT example from the manual on I-SAPT0/jun-cc-pVDZ between the two OH groups in a 2,4-pentanediol molecule. However, my job crashed with following error:

Traceback (most recent call last):
File “/software/psi4/1.3/python-3.7/bin/psi4”, line 287, in
File “”, line 48, in
File “/software/psi4/1.3/python-3.7/lib//python3.7/site-packages/psi4/driver/molutil.py”, line 250, in geometry
geom, enable_qm=True, missing_enabled_return_qm=‘minimal’, enable_efp=True, missing_enabled_return_efp=‘none’)
File “/software/psi4/1.3/python-3.7/lib//python3.7/site-packages/qcelemental/molparse/from_string.py”, line 260, in from_string
File “/software/psi4/1.3/python-3.7/lib//python3.7/site-packages/qcelemental/molparse/from_arrays.py”, line 123, in from_input_arrays
File “/software/psi4/1.3/python-3.7/lib//python3.7/site-packages/qcelemental/molparse/from_arrays.py”, line 381, in from_arrays
File “/software/psi4/1.3/python-3.7/lib//python3.7/site-packages/qcelemental/molparse/chgmult.py”, line 484, in validate_and_fill_chgmult
c_final, fc_final, m_final, fm_final = reconcile(cgmp_exact_c, cgmp_exact_fc, cgmp_exact_m, cgmp_exact_fm)
File “/software/psi4/1.3/python-3.7/lib//python3.7/site-packages/qcelemental/molparse/chgmult.py”, line 474, in reconcile
raise ValidationError(err)

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

Printing out the relevant lines from the Psithon --> Python processed input file:
H 2.03535 3.99216 1.53635
H 2.02481 4.96785 0.07455
symmetry c1
–> “”",“mol”)
core.set_global_option(“BASIS”, “jun-cc-pvdz”)
core.set_global_option(“SCF_TYPE”, “df”)
core.set_global_option(“GUESS”, “sad”)
core.set_global_option(“FREEZE_CORE”, “true”)

Psi4 stopped on: Friday, 05 April 2019 09:42AM
Psi4 wall time for execution: 0:00:00.08

To me it seems as a problem in reading charge of the multiplicity of the molecule. When I try to run FI-SAPT example with two phenols, everything is fine. Is there a bug in the case of three fragments or is it only some change in the molecule specification which is not updated in the manual?

Thank you very much for the advice!


This seems to be a known limitation of the ‘new’ way of handling molecules/fragments.

There is a note just above the example input:

Please have a look at both i-sapt examples.