Re: CCL:fractional nuclear charge

From: Rene Fournier - 1999-07-07 <renef!at!yorku.ca>
Subject: Re: CCL:fractional nuclear charge
Date: Thu, 6 Nov 2003 10:17:03 -0500 (EST)

Hello Steve, Mikael, Everyone:
 As Mikael wrote, with fractional nuclear charges...
 >The basis set might be a problem, as they are optimised for specific
 >atoms.
    I totally agree.
 >I guess one way of getting around this would be to use a combined
 >basis set for the metamorphosing atom, consisting of the BS of both
 >the initial (O) and the final (N) atoms.
 I am afraid this might give problems with near linear dependencies,
 maybe inacurrate inversion of the overlap matrix.  A way around that
 problem would be to define an "intermediate basis set" by
 interpolating
 two (preferably large) O and N basis sets that have the same contraction
 pattern, ie, a "metamorphosing basis set for a metamorphosing atom".
 I have a related question: does anyone know of a program that would
 allow to use positive charge distributions other than a set of point
 charges, for instance, a set of spherical Gaussian distributions.
 -- Rene
 ------------------------------------------
 |   Rene Fournier, Associate Professor   |
 |   Chemistry Dpt, York University       |
 |   4700 Keele Street,  Toronto          |
 |   Ontario, CANADA   M3J 1P3            |
 ------------------------------------------