evangelistalab / ocdft Goto Github PK

The following input file is supposed to run a CDFT calculation on a parallel dimer of ethene and fluorinated ethene. However when I run this input, the cdft multipliers (Vc) are all zero and don't appear to change at all. In fact I tried manually feeding the converged multipliers from an NWchem CDFT calculation and it still didn't optimize the constraints.

import cdft

molecule {
0 1
C 0.000000 0.000000 0.663493
C 0.000000 0.000000 -0.663493
H 0.000000 0.922503 1.236446
H 0.000000 -0.922503 1.236446
H 0.000000 -0.922503 -1.236446
H 0.000000 0.922503 -1.236446
"--" #Remove quotes around this before running
C 4.00 0.000000 0.660905
C 4.00 0.000000 -0.660905
F 4.00 1.103866 1.386732
F 4.00 -1.103866 1.386732
F 4.00 -1.103866 -1.386732
F 4.00 1.103866 -1.386732
}

set {
basis 6-31++G*
scf_type df
df_basis_scf cc-pvtz-jkfit
guess sad
reference uks
maxiter 200
e_convergence 8
d_convergence 6
molden_write true
}

set cdft {
charge [1.0,-1.0]
spin [1.0,-1.0]
}

set dft_functional b3lyp

energy('cdft')

For reference, the ground state energy using unconstrained DFT is -554.130239901998 hartree. According to an NWChem CDFT calculation, the total energy of the constrained system should be -553.777990832392 hartree with the following converged CDFT multipliers:

  1        0.1418185737
  2       -0.1117875392
  3       -0.1418185737
  4        0.1117875392

@fevangelista

Hello, I am interested in using this code for CDFT calculations with Psi4. I haven't tried compiling yet, but I notice that the CDFT function is commented out in main.cc. Is this part of the code functional or is this plugin more designed for performing OCDFT calculations?