Jr. Rabinowitz et Sb. Little, COMPARISON OF QUANTUM-MECHANICAL METHODS TO COMPUTE THE BIOLOGICALLY RELEVANT REACTIVITIES OF CYCLOPENTA-POLYCYCLIC AROMATIC-HYDROCARBONS, International journal of quantum chemistry, 52(3), 1994, pp. 681-691
In computational studies to understand the interaction of polycyclic a
romatic hydrocarbons (PAHs) with biomolecular systems, the semiempiric
al method AM1 has been used previously to determine the geometry of th
e PAH and its metabolites and relevant intermediates. A number of stud
ies have shown that AM1 provides geometries for parent PAHs that are a
cceptably close to experimentally determined structures. However, many
of the properties that determine the manner by which PAHs interact wi
th biological nucleophiles depend on the structure of metabolites and
reactive intermediates where less experimental information is availabl
e. In a previous study, we used AM1 to obtain the molecular geometries
of reactive intermediates of cyclopenta-PAHs (cPAHs) and then used si
ngle-point Hartree-Fock calculations, with the Gaussian 3-21g basis se
t, to obtain molecular energies and charge distributions, in order to
predict the direction of epoxide ring opening. Recent advances in the
availability of computational hardware and software have provided othe
r, more rigorous, methods for approaching this problem. In this study,
we used Hartree-Fock methods in the Gaussian series of programs emplo
ying the 3-21 g and 6-31 g basis sets and the local density functiona
l method Dmol to obtain molecular geometries, energies, and charge dis
tributions of the epoxides and the two potential hydroxycarbocations t
hat could result from protonated ring opening, for a series of cPAHs.
We have also performed the same calculations with AMSOL/SM2, a semiemp
irical method that adds the effect of the aqueous environment to the A
M1 Hamiltonian. The division of the cPAHs into classes is not altered
by these more rigorous calculations. The inclusion of water in the Ham
iltonian has a greater effect on the results than using the ab initio
methods to obtain the structure. (c) 1994 John Wiley & Sons, Inc.