exotic molecules or reactions by semiempirical calculations would be expected to
be unreliable for lack of appropriate parameterization. Semiempirical methods are
widely used in industry as an aid to the design of drugs and materials, and are very
likely employed in preliminary exploration of projects for which only the later, ab
initio results, ever see the light of publication.
To conclude: calculations from first principles are notnecessarilyto be preferred
to semiempirical ones, although for novel molecules and reactions ab initio-type
methods are more to be trusted.
References
- For a good, fairly compact account of molecular mechanics see Levine IN (2000) Quantum
chemistry, 5th edn. Prentice Hall, Engelwood Cliffs, NJ, pp 664–680 - (a) Dewar MJS (1975) Science 187:1037. (b) Halgren TA, Kleier DA, Lipscomb WN Science
190:591. (c) Dewar MJS (1975) Science 190:591
Chapter 1, Harder Questions, Answers
Q7
Both experiments and calculations can give wrong answers. Why then should
experiment have the last word?
This is a highly “philosophical” question, but we will try to answer it in a
practical way, relevant to our work as scientists.
First, we should note that in practice experiment does not automatically trump
calculations: calculations which are considered to be reliable have been used to
correct experimental results – or rather experimental claims, in contrast to “con-
firmed” experiments. Examples of this are the (likely) correction of dubious bond
energies [1] and heats of formation [2]. So the interesting question is, why should
“confirmed” experiments take precedence in credibility to calculations? Remove
the quotation marks and the question almost answers itself: as scientific realists
[3] we believe that a good experiment reflects a reality of nature; a calculation, on
the other hand, is a kind of model of nature, possibly subject to revision.
References
- Fattahi A, Lis L, Tian Z, Kass SR (2006) Angew Chem Int Ed Engl 45:4984
- Ventura ON, Segovia M (2005) Chem Phys Lett 403:378
- Leplin J (1997) A novel defence of scientific realism. Oxford University Press, Oxford
Answers 591