John Pople

John Pople: The Man Who Taught Computers Chemistry

Before John Pople, a chemist who wanted to know how a molecule would behave ran an experiment. After him, a chemist could type the molecule into a computer and get an answer that was often good enough to publish. Pople spent five decades converting quantum mechanics from an elegant theory that could be solved exactly for the hydrogen atom into working software that could be pointed at almost anything. In 1998 the Nobel Committee gave him half the Chemistry prize — for a body of work he insisted, to the end, was really mathematics.

Somerset to Trinity

Pople was born on 31 October 1925 in Burnham-on-Sea, Somerset. He attended Bristol Grammar School, and in 1943 won a scholarship to Trinity College, Cambridge, taking his BA in 1946. The war interrupted the smooth academic line: between 1945 and 1947 he worked at the Bristol Aeroplane Company. He returned to Cambridge and completed a PhD in mathematics in 1951 under John Lennard-Jones, writing on lone pair electrons — those valence electrons that sit unshared on an atom and quietly determine much of a molecule's shape and reactivity.

His doctoral work also produced a statistical-mechanical treatment of water that stood, by later assessment, as "the standard for many years." It was a characteristic early signal: Pople chose problems where a mathematical structure could be imposed on messy chemical reality.

Cambridge, the NPL, and the Move West

A research fellowship at Trinity followed his doctorate, and from 1954 he lectured in Cambridge's mathematics faculty. In 1958 he moved to London to head the basic physics division of the National Physical Laboratory. The following year he co-authored, with W. G. Schneider and H. J. Bernstein, *High Resolution Nuclear Magnetic Resonance* — a textbook that arrived precisely as NMR was becoming the instrument that would define modern structural chemistry.

Then came the decisive break. A 1961–62 sabbatical at Carnegie Mellon University in Pittsburgh persuaded him where the future of computational work lay, and in 1964 he moved there permanently. He remained in the United States for the rest of his life while keeping his British citizenship. In 1993 he moved again, to Northwestern University, as Trustees Professor of Chemistry, a post he held until his death.

Making Molecules Computable

Quantum mechanics in principle describes every chemical bond. In practice the equations become unsolvable the moment a molecule contains more than a couple of electrons. Pople's life's work was the systematic construction of approximations — defensible, testable simplifications that made the intractable tractable.

He developed the Pariser–Parr–Pople method for pi electron systems, the delocalized electrons that govern the behaviour of aromatic and conjugated molecules. He then pushed into three dimensions with CNDO in 1965 and the INDO approximations that followed, each trading a measured amount of rigour for an enormous gain in what could actually be calculated.

His more consequential move was toward *ab initio* methods — calculations built from first principles rather than fitted to experiment. Working with Slater-type and then Gaussian orbitals, he became the driving force behind the Gaussian suite of programs, co-authoring Gaussian 70, the first release. Gaussian did something no single paper could: it put quantum chemistry in the hands of chemists who were not quantum theorists. Alongside it he introduced the concept of "model chemistry" — the insistence that a computational method be defined precisely enough to be tested systematically against reality rather than tuned case by case. His group later produced the Gaussian-1 and Gaussian-2 composite methods, which stitched several calculations together to reach accuracies approaching experimental measurement.

The Gaussian Rupture

The software that carried his name also produced the ugliest episode of his career. Pople departed from Gaussian in 1991, and in the aftermath prominent scientists — Pople himself among them — were banned from using the program. The affair generated considerable controversy within the quantum chemistry community, a field small enough that such a schism was felt everywhere. Pople's response was to build a rival: after leaving, he developed the Q-Chem computational chemistry program.

Honours

The recognitions began early with the Mayhew Prize in 1948 and ran the length of his career: Fellow of the Royal Society in 1961, the Irving Langmuir Award in 1970, the Davy Medal in 1988, the Wolf Prize in Chemistry in 1992. In 1998 came the Nobel Prize in Chemistry, shared with Walter Kohn, for computational methods in quantum chemistry — an unusual award in that it honoured not a discovery about nature but a way of interrogating it. The Copley Medal followed in 2002 and a knighthood, as Knight Commander of the Order of the British Empire, in 2003.

The Private Man

Pople married Joy Bowers in 1952; she died of cancer in 2002. He identified as a Christian. He also insisted throughout his life that he was more a mathematician than a chemist — a self-assessment theoretical chemists have politely declined to accept, regarding him instead as among the most important figures their field has produced.

Death and Afterlife

He died of liver cancer in Chicago on 15 March 2004, survived by his daughter Hilary and his sons Adrian, Mark and Andrew. In 2009 his family donated his Nobel medal to Carnegie Mellon University, where the crucial work had been done. Bristol Grammar School named an IT room and a scholarship after him; the Pittsburgh Supercomputing Center named a supercomputer in his honour — which is, for a man who taught machines to do chemistry, exactly the right monument.