|
From: Toon V. <Too...@UG...> - 2006-07-02 12:15:44
|
Hi,
This might be a bizarre question, but I am interested in a simple way to
calculate e.g. the HF energy/gradient/whatever of a wavefunction
optimized with DFT/B3LYP. My interest is not restricted to these two
levels of theory, but any combination of one-body wavefunction methods.
I currently tried a simple approach, but ran into a few difficulties:
* first calculate a the wavefunction with method A, e.g. with water1.in:
% Single point calculation
mpqc:(
mole<CLKS>:(
basis<GaussianBasisSet>:(
molecule = $:molecule
name = "3-21G*"
)
molecule = $:molecule
functional<StdDenFunctional>:(
name = "B3LYP"
)
guess_wavefunction<CLHF>:(
total_charge = 0
molecule = $:molecule
basis<GaussianBasisSet>:(
molecule = $:molecule
name = "STO-3G"
)
)
total_charge = 0
)
)
molecule<Molecule>:(
geometry = [
[ -0.0073905643625 0.0261284427758 0.00708466177419 ]
[ 1.95058386069 -0.0137957715056 0.0492585954974 ]
[ -0.0469159377498 1.84102313597 0.0546961383553 ]
]
atoms = [ "O" "H" "H" ]
)
* Then run a second calculation with method B, using this input file:
% Single point calculation
mpqc:(
do_gradient = "yes"
mole<CLKS>:(
maxiter=0
basis<GaussianBasisSet>:(
molecule = $:molecule
name = "3-21G*"
)
molecule = $:molecule
functional<StdDenFunctional>:(
name = "BP86"
)
guess_wavefunction="water1.wfn"
total_charge = 0
)
)
molecule<Molecule>:(
geometry = [
[ -0.0073905643625 0.0261284427758 0.00708466177419 ]
[ 1.95058386069 -0.0137957715056 0.0492585954974 ]
[ -0.0469159377498 1.84102313597 0.0546961383553 ]
]
atoms = [ "O" "H" "H" ]
)
The maxiter=0 option does not work (tested with a recent cvs checkout).
This is no surprise since not too many people actually need this. Is
there a proper alternative for this trick? Any help will be appreciated.
best regards,
Toon
|