Printing molecular orbitals in input format

            This form allows molecular orbitals read from a file to be saved as an initial guess for new calculation. For example, with this utility one can use Gaussian computed MOs as an initial guess for GAMESS(US) job. If the orbitals contain spherical functions (5d, 7f ), they can be either converted into Cartesian functions (6d, 10f ) or printed as spherical (when writing orbitals in GAMESS(US) input format, such orbitals can only be converted to Cartesian). Chemcraft takes into account the fact that the sequence of F and G functions is not the same in different quantum chemistry packages.

            The checkboxes "Print atomic coordinates", "Print the basis set as non-standard" enable printing of molecular geometry and basis set parameters together with eigenvectors. If these options are disabled, you should ensure that the molecular geometry, including axes orientation, and the basis set are the same in the file from which the orbitals have been read and the input file where the eigenvectors should be written to. Note that this utility does not support printing of ECPs, so they must be specified in the input file manually if used in the computation.

            By default, the number of MOs to be printed is equal to the number of occupied orbitals in the opened file. Usually this value shouldn't be altered.

            Notes for GAMESS(US) users:

            When eigenvectors are printed in the GAMESS input file as cards, usually only C1 point group should be specified, with COORD=UNIQUE in $CONTRL group (otherwise the orbitals can be read improperly because of principal axes change in the molecule).

            If you are performing a DFT computation in GAMESS-US, you should also include SWOFF=0.0 in $DFT group (otherwise the orbitals are firstly converged in pure HF). This is not necessary with PCGAMESS.

            Notes for Gaussian users:

            With Gaussian98 usually NOSYMM must be specified in the route section of input file. Otherwise the orbitals are re-oriented by Gaussian and can be read improperly. With Gaussian03 these problems usually do not occur, but in some cases NOSYMM must still be specified in input.

 

 

 

 

BACK