DF-MP2 produces error for non-covalent interaction

Hi there, I am trying to optimize a halogen bonded complex with DF-MP2 using aug-cc-pVTZ. The input file is provided below.

memory 24 gb
molecule {
   0 1
   F        0.00000        0.00000        2.26150
   Cl        0.00000        0.00000        0.69160
   --
   0 1
   C        0.00000        0.00000       -3.10840
   H       -0.85740        0.00000       -3.81000
   H        0.85740        0.00000       -3.81000
  units angstrom
  no_reorient
}
set {
basis aug-cc-pvtz
df_basis_mp2 aug-cc-pvtz-ri
}
set {
scf_type df
freeze_core true
guess sad
}
set {
geom_maxiter 150
full_hess_every 10
opt_coordinates both
print_trajectory_xyz_file true
final_geom_write true
}
optimize('MP2')

The optimized complex should be in C2v symmetry. But I didn’t mention it anywhere. While running the job, I am getting an error,

psi4.driver.p4util.exceptions.ValidationError: Cannot compute projection of different symmetries.

How to solve this problem? Thank you in advance…
I am using Psi4 1.3.2 release

It seems that during the displacements for the numerical MP2 hessian the point group changes from C2v to Cs (or C2) and then the SCF restart from the previous orbitals fails with above error message.

The symmetry is autodetected.
An immediate remedy is a) forcing C1 symmetry at an increased computation cost. (put symmetry c1 in molecule block) or b) avoid calculating the Hessian

I believe @jmisiewicz is an expert on the FINDIF issue.

1 Like

Hi,

I tried with aug-ccpvDz and got the same error message. I took thes last coordinates and optimized them with b3lyp. Then I changed to aug-cc-pvTz and optimized again with b3lyp:

F            0.000000000000    -0.000000000000     1.816287793032
CL           0.000000000000    -0.000000000000    -0.216881726081
C            0.000000000000    -0.000000000000    -1.841212590418
H           -0.933731416408    -0.000000000000    -2.395213148818
H            0.933731416408     0.000000000000    -2.395213148818

Further optimization with MP2/Aug-cc-pVTZ leads to

F            0.000000000000    -0.000000000000     1.772804495357
CL           0.000000000000    -0.000000000000    -0.205507096512
C            0.000000000000    -0.000000000000    -1.810531270997
H           -0.933577256852    -0.000000000000    -2.365356953842
H            0.933577256852     0.000000000000    -2.365356953842
1 Like

Thank You.
The geometry seems okay.
However, F-Cl distance is 1.978 A, while C-Cl is 1.605 A. If it was just the opposite, it would have been better.
The singlet C always binds more closely with Halogen/Chalcogen atoms. I am trying to find the reason for it, but stuck there.
Thank you Dr. For your kind help.

Holger is correct. I’ll fix this in 1.5.