SAPT with ECP for Iodine atom

SAPT
1

Dear Developers,

I would like to perform SAPT calculation for molecular system which contains iodine atom where the ECP potential is included in the basis set.
Here is my input:

memory 80 Gb

molecule C6F4I2_a_Dimer {
0 1
C 3.95278 8.77906 17.27000
C 3.95370 8.66769 15.87621
C 4.69624 7.66256 15.27548
C 5.45289 6.78264 16.03403
C 5.45999 6.90438 17.40504
C 4.70962 7.89105 18.01259
I 2.84514 10.21948 18.29785
I 2.85283 9.91681 14.62401
F 4.69930 7.50619 13.95030
F 6.17155 5.82591 15.44708
F 6.18941 6.05798 18.13631
F 4.72541 7.94933 19.35386
–
0 1
C -0.68802 11.66405 16.83132
C -0.68894 11.55269 18.22511
C -1.43148 10.54756 18.82584
C -2.18813 9.66764 18.06729
C -2.19523 9.78938 16.69628
C -1.44486 10.77605 16.08873
I 0.41962 13.10448 15.80347
I 0.41193 12.80181 19.47731
F -1.43454 10.39119 20.15102
F -2.90679 8.71091 18.65424
F -2.92466 8.94298 15.96501
F -1.46065 10.83433 14.74746
units angstrom
}

set globals {
basis def2-tzvp
df_basis_scf def2-tzvp-jkfit
df_basis_sapt def2-tzvp-ri
df_basis_elst def2-tzvp-jkfit
scf_type df
}
energy(‘sapt2+(3)-ct’)

And the error message is:

Estimated memory usage: 61843.4 MB

Natural Orbital Cutoff:   1.000E-06
Disp(T3) Truncation:            Yes
CCD (vv|vv) Truncation:         Yes
MBPT T2 Truncation:             Yes

Traceback (most recent call last):
File “/scratch1/QC_Programe/psi4_1.3.2/bin/psi4”, line 287, in
exec(content)
File “”, line 62, in
File “/scratch1/QC_Programe/psi4_1.3.2/lib//python3.6/site-packages/psi4/driver/driver.py”, line 556, in energy
wfn = procedures[‘energy’][lowername](lowername, molecule=molecule, **kwargs)
File “/scratch1/QC_Programe/psi4_1.3.2/lib//python3.6/site-packages/psi4/driver/procrouting/proc.py”, line 3551, in run_sapt_ct
e_sapt = core.sapt(dimer_wfn, monomerA_wfn, monomerB_wfn)

RuntimeError:
Fatal Error: PSIO Error
Error occurred in file: /scratch/psilocaluser/conda-builds/psi4-multiout_1557940846948/work/psi4/src/psi4/libpsio/error.cc on line: 128
The most recent 5 function calls were:

psi::PSIO::rw(unsigned long, char*, psi::psio_address, unsigned long, int)
psi::PSIO::write_entry(unsigned long, char const*, char*, unsigned long)

Printing out the relevant lines from the Psithon --> Python processed input file:
core.set_global_option(“DF_BASIS_SAPT”, “def2-tzvp-ri”)

core.set_global_option("DF_BASIS_ELST", "def2-tzvp-jkfit")

core.set_global_option("SCF_TYPE", "df")

–> energy(‘sapt2+(3)-ct’)

Any help is welcome.

Thanks in advance
Attila

2

An error like that normally means that you ran out of memory. The memory estimate is quite far off… I’ll pass this along to the SAPT developers. In the meantime, the only recommendations I have are to try a less expensive SAPT variant or more memory.

3

I was just testing this and agree with @jmisiewicz. The below runs just fine and contains all the likely tripping points – iodine, ct, higher-order SAPT. Do you really need the components breakdown from SAPT2+(3)? SAPT0 generally gives good qualitative components, while for overall interaction energy, CCSD(T) will be more accurate than SAPT2+(3).

memory 30 Gb

molecule C6F4I2_a_Dimer {
0 1
H 0.0 0.0 0.0
I 1.5 0.0 0.0
--
H 1.5 3.0 0.0
I 0.0 3.0 0.0
units angstrom
}

set globals {
basis def2-tzvp
df_basis_scf def2-tzvp-jkfit
df_basis_sapt def2-tzvp-ri
df_basis_elst def2-tzvp-jkfit
scf_type df
}

energy('sapt2+(3)-ct')
4

Dear jmisiewicz and loriab

Thank you very much for your answers.
Finally, I solved my problem by setting the proper scratch directory instead of the default /tmp one.
It was a simple storage problem and I apologize for my disregard.
Of course, I already have the CCSD(T) energy values but I’m also interested in different interaction energy type components, since, I thing that iodine has a bit different behavior than smaller halogen atoms.
Maybe, it would be useful to include in the installation script a small answer about the setting of the scratch directory.
Thank you again for everything.
With best regards
Attila