hyperchem for heterocycle mos?

[friedman #*at*# tammy.harvard.edu (Dawn Friedman)]

 Mel Schriver asks (specifically):
 >
 >What it boils down to for me is wether
 >I can use AM1 or PM3 to calculate the orbitals of a heterocycle that
 >contains S and/or P.
 >
   Short answer:  yes.
   In terms of orbital energies, you will not get anything like the
 absolute energies you get with ab initio methods.  But your relative
 energies will be very useful.  We have just calculated HOMO-LUMO energy
 gaps for an extensive series of heterocycles and heterocyclic oligomers
 containing S, using PM3 in Hyperchem (and in G92) as well as ab initio
 calculations at STO-3G* and 3-21G*.  The absolute values of the HOMO-LUMO
 gaps are very different at PM3, but the ordering is the same, and the
 correlation with an experimental variable is almost identical.
   So I would say that relative orbital energies of related compounds at
 PM3 (PM3 because of the sulfur parametrization) can be valuable.
   As to whether PM3 orbitals look like STO-3G* or 3-21G* orbitals --
 in a general sense, yes; that is, trends are reproduced.  But for more
 details, well, I'll be looking into it myself...
   -- Dawn  friedman #*at*# tammy.harvard.edu