Page 21 - Strategies and Applications in Quantum Chemistry From Molecular Astrophysics to Molecular Engineer

6 J. TOMASI
quantum chemists, and also the stimulus of our discipline on the progress of hardware is
not negligible, as the example of Clementi tells.

Efficient computational algorithms for "in-depth calculations" produces, as output -
apparently this is a truism - accurate calculations. Let me consider now more in detail what
this means.

At the basis there is the tacit assumption of a reductionistic ideal. Quantum mechanics in
the current version is the correct theory, and the process of extracting from the whole
universe the molecule subjected to accurate calculations does not create problems. I do not
object this assumption, being however aware that there are objections, mainly for the
process of abstraction of one molecule from the whole universe ( see, e.g. Primas [13])
and for the definition of an isolated molecule (see, e.g. Wolley, [14],Claverie [15],
Sutcliffe[16]).
The material model is just a bit of matter - a molecule -, all the physical interactions are in
principle considered (even if some terms are discarded in actual calculations), the
modelization is thus reduced to the mathematical part. In addition, the report has the
characteristics of an explanation. Making reference to a celebrated sentence opining the
textbook on Quantum Chemistry by Eyring, Walter, Kimball [17]: "In so far as quantum
mechanics is correct, chemical questions are problems in applied mathemathics"; it may be
said that this program is a realization of that sentence.
This research program is far from being devoid of practical interest. Numerous are the
problems in which the interpretation of experimental evidence is dubious. ( Theoreticians
often forget that rarely experimental evidence is directly amenable to the properties of more
general interest for the progress of our knowledge of the material work; at the molecular
level at least theory and computations have the advantage of getting directly the property of
interest.) In case of doubt, as in the choice between two different values of a molecular
parameter, both fitting the experimental data, the computed value often has univocally
provided the right answer.
In many other cases the experimental approach is unable to give the desired answer,
because the material system is not available in laboratory, as it is the case for many
astrophysical problems, or because the experiment is too complex and delicate. I am unable
to find now the exact source of the remark performed some years ago by a well known
experimentalist that the determination of the octupole moment of a molecule requires a stay
of the molecule in the measuring device for a time of the order of a week, while a computer
gives an equivalent answer within few hours. I am not expert of the progresses in
measuring octupole moments, but surely the quality of the quantum mechanical results in
the meantime greatly ameliorated, the computer time reduced, the size of the molecules for
which this calculation is possible is noticeably enlarged (We may also add that the use of
this kind of program may be now left to a medium-grade technician, while the experimental
determination requires a good technical skill. This remark is not essential here, but has its
weight in the general economy of the research).
One of the fields in which the competition of ab-initio methods is evident is that of
molecular vibrations. The experimental technique is relatively easy - transient species apart-
but the ab-initio methods give now results of comparable quality and a wealth of additional
information [18]. It is reasonable to forecast that the number of experimental measurements
which will find in accurate ab-initio calculations a serious competitor will increase.