Page 24 - Strategies and Applications in Quantum Chemistry From Molecular Astrophysics to Molecular Engineer

QUANTUM CHEMISTRY: THE NEW FRONTIERS 9
others which, according to the Primas definition [13] are graceless. These graceless
quantities are defined and selected with the scope of "understanding", i.e. in our
terminology with the scope of giving an interpretation of the chemical phenomena.
The "nature of the chemical bond" , the "chemical group effects" are examples of
"concepts" accepted by group II as objects of theoretical investigation. To perform these
studies it is allowed to introduce other "concepts" and "quantities" which have a
questionable status in the formal theory.

There is a large number of concepts and quantities of this kind used in actual
investigations, and a large rate of increase of new formulations. This abundance of
interpretative tools could lead somebody to suspect that we have lost any control on the
growth and on the use of of these instruments and that interpretation in chemistry is
becoming an exercise in which is possible to reach the conclusions one desires by an
appropriate selection of the tools.
This is not my opinion, and I will try to explain why.

Interpretation, as it has been already said, is not univocal. There is competition among
different interpretations, and the concept of "generality" of an interpretation (i.e. its range
of applicability) should be - and in fact is - an important criterion to eliminate ad hoc
descriptions. In other words a good chemical concept must be robust (the adjective is
taken again from the Primas book [13]). Gracelessness and robustness must be balanced.
As example we may consider the natural orbitals. These quantities have been proposed by
group I: they have "grace" and little effectiveness in the interpretation. Their use in the
Weinhold's formulation of natural bond analysis (NBOA) [20] makes them graceless but
effective; it rests to verify if this formulation is robust enough.
The definition of "concepts" must be accompanied by explicit recipes for computing them
is actual cases. There is no more space in theoretical chemistry for "driving forces",
"effects', etc. not accompanied by specific rules for their quantification. The impact of a
new "concept' will be greater if the rules of quantifications are not restricted to ad hoc
methods, but related to methods of general use in molecular quantum mechanics. A
concept based exclusively on some specific features of a given method, e.g. the extended
Hückel method, is less robust than a concurrent concept which may be quantified also
using other levels of the theory.

The "chemical concepts" represent a part of the model and must share with the entire model
other requirements, in particular simplicity, falsicability, and agreement with the general
laws of physics [6]. These additional criteria make possible to keep under control the
growth of methodological proposals.
The elaboration of "concepts" often requires the partition of the molecule into smaller
subunits. This partition is not supported by formal theories, and it is thus at a good extent
arbitrary. The consideration of the above mentioned criteria introduces strong limitations in
the choice of submolecular units. In fact there are only three basic choices: the constituent
atoms, the molecular orbitals and the partition of the charge distribution into localized
units. Each choice presents advantages and disadvantages which is not convenient to
analyze here.
The selection of a type of basic subunit is the first step in the elaboration of interpretative
tools. An analysis of this work, which represent the essence of the innovative activity of
group II is not possible here. It sufficient to remark that during this process of elaboration
there has been important "admixtures" of concepts having their origin in different choices
of the basic subunits.