|
Internet Electronic Journal of Molecular Design - IEJMD, ISSN 1538-6414, CODEN IEJMAT
| ABSTRACT - Internet Electron. J. Mol. Des. July 2006, Volume 5, Number 7, 364-375 |
The Evolution of the Valence Delta in Molecular Connectivity Theory
Lionello Pogliani
Internet Electron. J. Mol. Des. 2006, 5, 364-375
|
Abstract:
The evolution of a fundamental parameter in Kier and Hall molecular
connectivity theory, the valence delta δv, is followed from its
beginnings till its most recent definitions. This parameter, which was
initially based on quantum concepts (atomic number, and valence
electrons), is now solely based on concepts belonging to general and
complete graphs, like multiple edges, and self-connections of
pseudographs, and order and regularity of complete graphs.
Furthermore, the hydrogen content of a molecule, which was implied
in hydrogen suppressed chemical graphs, is now directly encoded into
the valence delta. These two new features of the valence delta allow to
graph differentiate among atoms of any type. The model quality of the
newly defined valence delta is tested with three properties of three
different classes of compounds: side-chain molecular volume, and
isoelectric point of amino acids, and boiling points of amines plus
boiling points of alcohols. The model of the composite class [amines +
alcohols], which seems to behave as a new property underlines the
peculiar character of the new valence delta. A new valence delta
number, which includes hydrogen perturbation, is defined. The defined
hydrogen perturbation makes no use of any new graph concept
because it is the ratio between two different valence delta numbers.
The ratio is always smaller than one and can be fine-tuned thanks to an
exponential parameter. A model of three properties with and without
hydrogen perturbation is presented. The model, which is achieved with
different values for the exponential parameter, underlines the
advantages and characteristics of this kind of perturbation. The
hydrogen perturbation throughout the present model study is not only
property-dependent, but also N-dependent, i.e., dependent on the
number of studied compounds.
|