Qualitative Thinking in the Age of Modern Computational Chemistry, or What Lionel Salem Knows

The achievements of modern computational chemistry are astounding. It is reasonable today to handle billions of configurations, and to achieve chemical accuracy, kilocalories say, in calculating binding energies and geometries, in ground and transition states of reasonably complex molecules. There is no question that the enterprise of computational theoretical chemistry is successful. Now Lionel Salem and I grew up and developed scientifically in the climate of the very same computer which made all this possible. Russ Pitzer taught me to punch cards; I still miss the sound of the key punch. The extended Hückel method, which several of us developed in the Lipscomb group, would have been impossible without modern computers. But I took a different turn, moving from being a calculator in the framework of semiempirical theory, to being an explainer, the builder of simple molecular orbital models. I was and am still doing calculations, but my abiding interest is in the construction of explanations. And also in thinking up moderately unreasonable things for experimentalists to try. In existing as a scientist, meaning that my work was of continuing interest to other chemists, I was helped in that I moved into whatever part of chemistry I did, just a little ahead of the heavy guns of computational chemistry. So I switched from organic to inorganic molecules just when organic molecules became reasonably calculable. Recently I’ve been less fortunate, for when I moved to solids and surfaces I came back into heavy fire—physicists had been doing calculations on these materials for a long time. And they were (are) hardly likely to believe that one-electron calculations and a chemical viewpoint are of value. I want to make some observations on computational quantum chemistry, perforce influenced by my prejudices. Given the advances in the field, any molecule I can calculate (without geometrical optimization), with the simplest extended Hückel approximation, can be done so much better by most computational chemists. So why don’t I feel threatened; why is there a role for people of my ilk? Or for Lionel. Actually, I do feel threatened and bypassed! But that’s just an emotional reaction, and my aging figures in it.