The "birth-year" of computational chemistry?



I dug into this question about 12 years ago.  It depends of course on just
 what you call computational chemistry.  Does it have to be using a
 computer?  I would argue that should not be a requirement.  Rather the
 real underlying field is the construction of a numeric simulation that
 represents 'reality' in a context that permits testing concepts and
 thereby learning something.  Today we use computers in their current form,
 perhaps this will change in the decades to come.
 I point back to early spectroscopic work.  The earliest I think is work by
 Andrews, in the 20s.  Once citation I have is in Physical Reviews, v36,
 p531, 1930, Kettering, Shutts and Andrews (General Motors Corporation).
 "A representation of the Dynamic Properties of Molecules by Mechanical
 Models".  I think of this as the birth of molecular modeling.  They built
 molecular models of things like toluene, benzene, chlorobenzene using
 steel balls and calibrated springs.  They attached an oscillator and swept
 the frequency.  They then used a strobe light an found the standing
 waves.  In this way they mapped out a molecular mechanics model the IR and
 Raman spectra.  They were then able to draw conclusions about the correct
 molecular structure by doing the experiment on various bonding models of
 benzene by having springs attached in different arrangements.  I think
 they got it wrong in the end, but that sure looks like molecular mechanics
 with an analog computer to me.
 There is also the work by Urey and Bradley, whose names are still attached
 to a force field, that begins around 1931 with a paper in Physical Review,
 p1969, 1931 (apparently Bradley's Ph.D. thesis).  He discusses Andrew's
 papers and also a paper I never saw by Bjerrum in physik. Ges. Ber.  v16,
 p737, 1914, proposing those sorts of forces acting upon bonded atoms.  He
 of course also cites Slater (1931) and Pauling JACS,v53,p1367, 1931, as
 discussing the nature of the forces on a chemical bond.
 Still later, I put Frank Westheimer (with John Mayer) at the birth of
 computational chemistry using numerical methods.  He calculated an
 activation energy for racemization of 2,2'-dibromo-4,4'-dicarboxydiphenyl
 in J. Chemical Physics, v15, p252, 1947 (there is also an earlier paper
 v14, p733,1946).  He used an exponential function to describe the rates
 and pull out constants.
 Other names at that time are T.L. Hill, J. Chem. Phys. v14,465 (1946)
 and W. Gordy, J.Chem.Phys 14,305(1946).
 __
 M. Dominic Ryan   (610)-270-6529     SmithKline Beecham Pharmaceuticals
 Internet:  ryanmd[ AT ]mh.us.sbphrd.com     King of Prussia, PA