XI
Simpósio Brasileiro de Química Teórica
Caxambu,
18 a 21 Novembro de 2001
|
Rudolf
Kiralj (PQ) and Márcia M. C. Ferreira (PQ), Instituto de Química,
Universidade
Estadual
de Campinas, Campinas, SP, 13083-970, Brazil. marcia@iqm.unicamp.br
Linear polyacenes,
although being the simples aromatic hydrocarbons, are
known up to
the seventh
member of the series (heptacene).1
They are characterized by rapid
increase in reactivity
and decrease in stability as the chain
length increases. Such an
aromaticity is
called by Clar2 as
acene character; it includes meso-rings
in aromatic
substitution/addition,
and terminal rings in some other reactions
as the most reactive
sites. The
following reactivity of linear polyacenes
was studied in this work by semi-
empirical PM3:
aromatic meso-bromination, biochemical
and artificial ortho-oxidation,
total hydrogenation,
meso-additions (hydrogenation, photochemical
oxidation, addition
of bromine and maleic acid
anhydride) and meso-oxidation to quinones.
The geometries
of reactants and products
were optimized, and heats of formation of
reactions calculated
from heats of formation
and total energies of the molecules. Molecular descriptors as graph-
theoretical and
PM3 resonance energy, hardness and Dewar's
localization, para- and
ortholocalization numbers
were calculated. The heats of the reaction were
analyzed as a
function of chain length,
and the reactions with similar energy
profile were studied by
Principal Component
(PCA) and Hierarchical Cluster Analysis
(HCA). Partial Least
Squares Regression
(PLS) related heats of reaction to the molecular descriptors
and the
chain length.
Energy profiles for meso-ring reactions have
the same shape. They are
described well
in PCA by one Principal Component (over
90% of original information).
Both PCA
and HCA show that acene character
begins with anthracene, with high
similarity between
molecules having the same number of
different reactive sites. Both
analyses distiguish
anti- from syn- and
in-plane oriented reactions.
PLS models
(Q > 0.96, R
> 0.98) are superior to linear regression models (R > 0.75).
1 W. J. Bailey, C.-W. Liao,
J. Amer. Chem. Soc. 77 (1955) 992-994.
2 E. Clar: The Aromatic
Sextet, Wiley&Sons, New York, 1972.
(FAPESP