Chapter 6, Harder Questions, Answers
Q8
It has been said of SE methods: “They will never outlive their usefulness for
correlating properties across a series of molecules...I really doubt their predictive
value for a one-off calculation on a small molecule on the grounds that whatever
one is seeking to predict has probably already been included in with the para-
meters.” (Hinchcliffe A (1987) Ab initio determination of molecular properties.
Adam Hilger, Bristol, p. x). Do you agree with this? Why or why not? Compare the
above quotation with Dewar MJS (1992) A semiempirical life. American Chemical
Society, Washington, DC, pp 133–136.
First, a synopsis of Dewar pp. 133–136. Here are representative excerpts:
One of the criticisms commonly levelled at semiempirical methods is that they represent no
more than methods of interpolation and are useful only in areas and for compounds for
which they have been parameterized....The striking thing about ours is that they donot
merely reproduce the properties for which they were parameterized, nor are they confined
to molecules of the kind used in the parameterization. They reproduceallground-state
properties of molecules ofallkinds...Thus our procedures provide a very good representa-
tion of the way molecules behave...”
“...our work has led to a number of predictions that have been subsequently confirmed
by experiment.”
“Every procedure performs less well in some cases than in others. How serious each
error is depends on the chemical importance of the molecule in question.”
The statements above directly contradict the assertion that “...whatever one is
seeking to predict has probably already been included in with the parameters.”, with
the reservation that Hinchcliffe was presumably writing about 5 years before
Dewar. The references given by Dewar, and the experience of the many chemists
who use semiempirical methods (not only the Dewar-type ones) show that these are
notmerely “methods of interpolation”. It is however true that for accurate, reliable
information on the properties of a small molecule one would very likely resort to a
high-level ab initio or DFT calculation.
Chapter 6, Harder Questions, Answers
Q9
For common organic molecules Merck Molecular Force Field geometries are
nearly as good as MP2(fc)/6-31G geometries. For such molecules single point
MP2(fc)/6-31G calculations, which are quite fast, on the MMFF geometries,
should give energy differences comparable to those from MP2(fc)/6-31G//MP2
(fc)/6-31G calculations [geometry optimizations at the MP2(fc)/6-31G* level].
634 Answers