PC GAMESS-related discussion club

Alex Granovsky
gran@classic.chem.msu.su

Indeed, if you omit all terms but pure electrostatics,
this most likely results in Coulomb collapse of your system.
In any case, you will not be able to find the optimal geometry.
Thus, you need both the ab-initio-EFP and EFP-EFP Pauli repulsion terms.

Unfortunately, the PC GAMESS does not allow fully automated
calculation of these terms for arbitrary fragment.

There exists internally stored standard fragment for H2O molecule,
which can be used in calculations and you do not need to specify
any additional parameters for it.

However, for other solvents, you can obtain all the required
electrostatic terms using runtyp=makefp, but not the parametrization
of repulsion terms. Thus, you need to adjust some (initially guessed)
parameters of repulsion manually using a set of QC calculations like
Morokuma decomposition, etc...

As far as I remember, there should be some program available
which was developed and used by the authors of the EFP model
to fit repulsive terms for water and other molecules.


Best regards,
Alex Granovsky


On Sat Oct 11 '08, Fumihito Mohri wrote
---------------------------------------
>Dear  All
>Now, I am interested in the application of EFP method to solvent effect computation,
>since I think that solvent effect computation by EFP is more sophisticated than PCM.
> So, I read the description of $EFRAG in INPUT.pdf, and then I found the following
>explanation.
>        POSITION=OPTIMIZE Allows full optimization within the
>                           ab initio part, and optimization of
>                           the rotational and translational
>                           motions of each fragment. (default)

>In EFP method, each "fragment molecule" is treated as multi-poles (a set of point
> charges, one dipole or quadrupole etc). But, I consider with Earnshaw's theorem,
>in order to obtain the optimization of the rotational and translational motions,
>I should determine the form of exchange repulsions between fragment molecules and
>between fragment molecules and ab-initio molecules. Then, I have a question: Should
>I define the REPULSIVE POTENTIAL in $FRAGNAME and the inter-fragment repulsive potential in $FRGRPL?

> Best regards,
> Fumihito Mohri
>